From flyingdeckchair at googlemail.com  Wed Oct  1 04:59:46 2008
From: flyingdeckchair at googlemail.com (peter websdell)
Date: Wed, 1 Oct 2008 09:59:46 +0100
Subject: [SciPy-user] unusual fitting problem
In-Reply-To: <200809301240.54583.pgmdevlist@gmail.com>
References: <b4246c960809300936q325f2e75ubed1c90582a45025@mail.gmail.com>
	<200809301240.54583.pgmdevlist@gmail.com>
Message-ID: <b4246c960810010159g6f230198sd4c64b8eefca795c@mail.gmail.com>

Hello,
Thanks for the reply Pierre.

I may be misunderstanding you, but it seems to me that if I interpolate
between two values, the relationship will be linear. What I need is for it
to appear linear when plotted on a loglog scale.

I'm sure I'm not explaining this very well, owing to this not really being
my field, so I've attached an image of the kind of curve I'm attempting to
fit.

Please let me know if I'm just being a dunce.

Thanks again,
Pete.

2008/9/30 Pierre GM <pgmdevlist at gmail.com>

> On Tuesday 30 September 2008 12:36:00 peter websdell wrote:
> > Howdy gang,
> > I have an unusual fitting problem that has me totally stumped.
>
> Have you thought about modifying your data ?
> you want log(y) = a + b*log(x) ?
> Use Y=log(y) and X=log(X), fit a straight-line the standard way and you
> should
> get your parameters a & b.
> That should be easier than trying to fit y=exp(a+b*log(x)) = exp(a) * x**b
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081001/dd14feaf/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: curve1.JPG
Type: image/jpeg
Size: 70310 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081001/dd14feaf/attachment.jpe>

From massimo.sandal at unibo.it  Wed Oct  1 05:28:03 2008
From: massimo.sandal at unibo.it (massimo sandal)
Date: Wed, 01 Oct 2008 11:28:03 +0200
Subject: [SciPy-user] unusual fitting problem
In-Reply-To: <b4246c960810010159g6f230198sd4c64b8eefca795c@mail.gmail.com>
References: <b4246c960809300936q325f2e75ubed1c90582a45025@mail.gmail.com>	<200809301240.54583.pgmdevlist@gmail.com>
	<b4246c960810010159g6f230198sd4c64b8eefca795c@mail.gmail.com>
Message-ID: <48E342A3.1080701@unibo.it>

peter websdell wrote:
> Hello,
> 
> Thanks for the reply Pierre.
> 
> I may be misunderstanding you, but it seems to me that if I interpolate 
> between two values, the relationship will be linear. What I need is for 
> it to appear linear when plotted on a loglog scale. 

If I understand correctly: The relationship will be linear, but between 
the *logarithms*.

By using Y=log(y) and X=log(x) mathematically it's just having log/log 
paper. You are just linearizing the equations.

m.
-- 
Massimo Sandal , Ph.D.
University of Bologna
Department of Biochemistry "G.Moruzzi"

snail mail:
Via Irnerio 48, 40126 Bologna, Italy

email:
massimo.sandal at unibo.it

web:
http://www.biocfarm.unibo.it/samori/people/sandal.html

tel: +39-051-2094388
fax: +39-051-2094387
-------------- next part --------------
A non-text attachment was scrubbed...
Name: massimo_sandal.vcf
Type: text/x-vcard
Size: 274 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081001/d99d9b10/attachment.vcf>

From flyingdeckchair at googlemail.com  Wed Oct  1 05:36:01 2008
From: flyingdeckchair at googlemail.com (peter websdell)
Date: Wed, 1 Oct 2008 10:36:01 +0100
Subject: [SciPy-user] unusual fitting problem
In-Reply-To: <48E342A3.1080701@unibo.it>
References: <b4246c960809300936q325f2e75ubed1c90582a45025@mail.gmail.com>
	<200809301240.54583.pgmdevlist@gmail.com>
	<b4246c960810010159g6f230198sd4c64b8eefca795c@mail.gmail.com>
	<48E342A3.1080701@unibo.it>
Message-ID: <b4246c960810010236g5c3f69dcma51c12c205eaeeb0@mail.gmail.com>

Yes! Of course. Apologies for the stupid question.
Cheers,
Pete.

2008/10/1 massimo sandal <massimo.sandal at unibo.it>

> peter websdell wrote:
>
>> Hello,
>>
>> Thanks for the reply Pierre.
>>
>> I may be misunderstanding you, but it seems to me that if I interpolate
>> between two values, the relationship will be linear. What I need is for it
>> to appear linear when plotted on a loglog scale.
>>
>
> If I understand correctly: The relationship will be linear, but between the
> *logarithms*.
>
> By using Y=log(y) and X=log(x) mathematically it's just having log/log
> paper. You are just linearizing the equations.
>
> m.
> --
> Massimo Sandal , Ph.D.
> University of Bologna
> Department of Biochemistry "G.Moruzzi"
>
> snail mail:
> Via Irnerio 48, 40126 Bologna, Italy
>
> email:
> massimo.sandal at unibo.it
>
> web:
> http://www.biocfarm.unibo.it/samori/people/sandal.html
>
> tel: +39-051-2094388
> fax: +39-051-2094387
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081001/4383f922/attachment.html>

From stefan at sun.ac.za  Wed Oct  1 06:06:48 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Wed, 1 Oct 2008 12:06:48 +0200
Subject: [SciPy-user] sparse matrices -- slicing and fancy indexing for
	assignment
In-Reply-To: <d05265cb0809300901i2a20facam421af339cb8238b7@mail.gmail.com>
References: <4FBBE3B4-041E-4593-8ADA-1318BF43A853@mit.edu>
	<d05265cb0809300901i2a20facam421af339cb8238b7@mail.gmail.com>
Message-ID: <9457e7c80810010306j433fcc2cyb5f78b45bee9e150@mail.gmail.com>

2008/9/30 Nathan Bell <wnbell at gmail.com>:
> This is probably all you'll see in 0.7.0
>
> Note that assigning to a CSR/CSC matrix is in general a bad idea.  Any
> change to the sparsity structure of these formats requires O(nnz)
> operations, which basically means reconstructing the matrix from
> scratch.
>
> OTOH MATLAB lets you do it, so we will probably support it someday
> too.  If you submit a patch soon we might be able to integrate it by
> 0.7.0.

I was about to document this, but Nathan has already done the job:

http://projects.scipy.org/scipy/scipy/browser/trunk/scipy/sparse/info.py

Regards
St?fan


From philbinj at gmail.com  Wed Oct  1 08:59:44 2008
From: philbinj at gmail.com (James Philbin)
Date: Wed, 1 Oct 2008 13:59:44 +0100
Subject: [SciPy-user] r4763 won't compile, missing lsodar.pyf
Message-ID: <2b1c8c4f0810010559v6defd4bbsd510798cfd47a55c@mail.gmail.com>

Hi,

"python setup.py build"
errors with:
"target build/src.linux-x86_64-2.5/lsodarmodule.c does not exist:
   Assuming lsodarmodule.c was generated with "build_src --inplace" command."

And indeed, scipy/integrate/setup.py has:
"config.add_extension('lsodar',
                         sources=['lsodar.pyf'],
                         libraries=libs,
                         **newblas)"

But scipy/integrate doesn't contain lsodar.pyf?

Thanks,
James


From robince at gmail.com  Wed Oct  1 09:50:09 2008
From: robince at gmail.com (Robin)
Date: Wed, 1 Oct 2008 14:50:09 +0100
Subject: [SciPy-user] broadcasting elementwise on sparse matrix
Message-ID: <b14b627d0810010650q26130e48x789ae477a2429a6d@mail.gmail.com>

Hi,

Is there any way to get broadcasted element wise operations on sparse matrices?

I would like to do something like A * x if A was (n,n) array and x was
a (n,) array. I saw the sparse matrices have a .multiply method, but
it doesn't seem to broadcast in the same way as the standard
multiplication, ie As.multiply(x) gives an "inconsistent shapes"
error.

The best I have is As.multiply(tile(x,(n,1))) although this has the memory
overhead that is normally avoided with broadcasting. Is there an alternative?

Thanks

Robin


From travis at enthought.com  Wed Oct  1 10:36:54 2008
From: travis at enthought.com (Travis Vaught)
Date: Wed, 1 Oct 2008 09:36:54 -0500
Subject: [SciPy-user] Texas Python Regional Unconference Reminders
Message-ID: <0107962E-D762-497B-BCEC-24CBD78B381B@enthought.com>

Greetings,

The Texas Python Regional Unconference is coming up this weekend  
(October 4-5) and I wanted to send out some more details of the  
meeting.  The web page for the meeting is here:

http://www.scipy.org/TXUncon2008

The meeting is _absolutely free_, so please add yourself to the  
Attendees page if you're able to make it.  Also, if you're planning to  
attend, please send me the following information (to travis at enthought.com 
) so I can request wireless access for you during the meeting:

  - Full Name
  - Phone or email
  - Address
  - Affiliation

There are still opportunities to present your pet projects at the  
meeting, so feel free to sign up on the presentation schedule here:

http://www.scipy.org/TXUncon2008Schedule

For those who are in town Friday evening, we're planning to get  
together for a casual dinner in downtown Austin that night.  We'll  
meet at Enthought offices (http://www.enthought.com/contact/map-directions.php 
) and walk to a casual restaurant nearby.  Show up as early as 5:30pm  
and you can hang out and tour the Enthought offices--we'll head out to  
eat at 7:00pm sharp.

Best,

Travis




From stefan at sun.ac.za  Wed Oct  1 10:58:38 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Wed, 1 Oct 2008 16:58:38 +0200
Subject: [SciPy-user] broadcasting elementwise on sparse matrix
In-Reply-To: <b14b627d0810010650q26130e48x789ae477a2429a6d@mail.gmail.com>
References: <b14b627d0810010650q26130e48x789ae477a2429a6d@mail.gmail.com>
Message-ID: <9457e7c80810010758w4973a9e9v2fe1f238b5d85ccd@mail.gmail.com>

Hi Robin

2008/10/1 Robin <robince at gmail.com>:
> Is there any way to get broadcasted element wise operations on sparse matrices?
>
> I would like to do something like A * x if A was (n,n) array and x was
> a (n,) array. I saw the sparse matrices have a .multiply method, but
> it doesn't seem to broadcast in the same way as the standard
> multiplication, ie As.multiply(x) gives an "inconsistent shapes"
> error.
>
> The best I have is As.multiply(tile(x,(n,1))) although this has the memory
> overhead that is normally avoided with broadcasting. Is there an alternative?

You can broadcast over the indices, and then use them to individually
address the elements of the sparse matrix.  Not super-fast, but better
than nothing:

In [4]: a, b = np.broadcast_arrays([1,2,3], [[1,2,3], [4,5,6]])

In [5]: a
Out[5]:
array([[1, 2, 3],
       [1, 2, 3]])

In [6]: b
Out[6]:
array([[1, 2, 3],
       [4, 5, 6]])

Cheers
St?fan


From wnbell at gmail.com  Wed Oct  1 11:52:32 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Wed, 1 Oct 2008 11:52:32 -0400
Subject: [SciPy-user] broadcasting elementwise on sparse matrix
In-Reply-To: <b14b627d0810010650q26130e48x789ae477a2429a6d@mail.gmail.com>
References: <b14b627d0810010650q26130e48x789ae477a2429a6d@mail.gmail.com>
Message-ID: <d05265cb0810010852u2e3292d7weca59331f5647312@mail.gmail.com>

On Wed, Oct 1, 2008 at 9:50 AM, Robin <robince at gmail.com> wrote:
>
> Is there any way to get broadcasted element wise operations on sparse matrices?
>
> I would like to do something like A * x if A was (n,n) array and x was
> a (n,) array. I saw the sparse matrices have a .multiply method, but
> it doesn't seem to broadcast in the same way as the standard
> multiplication, ie As.multiply(x) gives an "inconsistent shapes"
> error.
>
> The best I have is As.multiply(tile(x,(n,1))) although this has the memory
> overhead that is normally avoided with broadcasting. Is there an alternative?
>

That particular operation is just scaling of the columns, right?  If
so, you can use:
>>> As * spdiags([x],[0], n,n,)

If As is already a CSR matrix, then you can cheat a little and do the
operation in place:
>>> As.data *= x[As.indices]

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From robince at gmail.com  Wed Oct  1 12:11:12 2008
From: robince at gmail.com (Robin)
Date: Wed, 1 Oct 2008 17:11:12 +0100
Subject: [SciPy-user] broadcasting elementwise on sparse matrix
In-Reply-To: <d05265cb0810010852u2e3292d7weca59331f5647312@mail.gmail.com>
References: <b14b627d0810010650q26130e48x789ae477a2429a6d@mail.gmail.com>
	<d05265cb0810010852u2e3292d7weca59331f5647312@mail.gmail.com>
Message-ID: <b14b627d0810010911q74cf30c8r49eee9f70f2c3385@mail.gmail.com>

On Wed, Oct 1, 2008 at 4:52 PM, Nathan Bell <wnbell at gmail.com> wrote:
> On Wed, Oct 1, 2008 at 9:50 AM, Robin <robince at gmail.com> wrote:
> That particular operation is just scaling of the columns, right?  If
> so, you can use:
>>>> As * spdiags([x],[0], n,n,)

Yes - I realised that and was using spdiags, which is a lot quicker.

> If As is already a CSR matrix, then you can cheat a little and do the
> operation in place:
>>>> As.data *= x[As.indices]

That's quite tricky... :)
It is a CSC matrix - not sure why, but I am saving and loading it from
Matlab files, and when I first started (a year or so again) it seemed
CSC was the most reliable (or at least I always got a CSC from .mat
files).
Presumably the same thing the same thing will work though.

Thanks very much,

Robin


From wnbell at gmail.com  Wed Oct  1 12:17:36 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Wed, 1 Oct 2008 12:17:36 -0400
Subject: [SciPy-user] broadcasting elementwise on sparse matrix
In-Reply-To: <b14b627d0810010911q74cf30c8r49eee9f70f2c3385@mail.gmail.com>
References: <b14b627d0810010650q26130e48x789ae477a2429a6d@mail.gmail.com>
	<d05265cb0810010852u2e3292d7weca59331f5647312@mail.gmail.com>
	<b14b627d0810010911q74cf30c8r49eee9f70f2c3385@mail.gmail.com>
Message-ID: <d05265cb0810010917l65d7535cr5f6f5d481c03424@mail.gmail.com>

On Wed, Oct 1, 2008 at 12:11 PM, Robin <robince at gmail.com> wrote:
>
> That's quite tricky... :)
> It is a CSC matrix - not sure why, but I am saving and loading it from
> Matlab files, and when I first started (a year or so again) it seemed
> CSC was the most reliable (or at least I always got a CSC from .mat
> files).
> Presumably the same thing the same thing will work though.
>

Yep, except it would scale rows.  I'd stick with spdiags() for safety.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From robince at gmail.com  Wed Oct  1 12:20:19 2008
From: robince at gmail.com (Robin)
Date: Wed, 1 Oct 2008 17:20:19 +0100
Subject: [SciPy-user] dot behaviour with sparse matrix
Message-ID: <b14b627d0810010920m4a994a1q23feb756397c66c3@mail.gmail.com>

Not sure if this is a bug or not but it was certainly unexpected
behaviour for me.

I found that dot(dense,dense) works as expected
(ie dimensions (8,255),(255,8) -> (8,8))
and that dot(sparse,sparse) also works that way, producing a sparse result.

But dot(dense,sparse) gives an output with the same dimensions as the
dense input. Not sure why this is happening and I found it a bit
confusing. Is this a bug or inteded behaviour?

Actually I was just going through an example and I notice that
dot(csc_matrix, dense_matrix) raises a NotImplemented type - so
perhaps the other cases should also:

With A, B arrays:
In [29]: A.shape
Out[29]: (3, 5)
In [30]: B.shape
Out[30]: (5, 3)
In [31]: dot(A,B).shape
Out[31]: (3, 3)
In [32]: dot(sparse.csc_matrix(A),sparse.csc_matrix(B)).shape
Out[32]: (3, 3)
In [33]: dot(A,sparse.csc_matrix(B)).shape
Out[33]: (3, 5)
In [34]: dot(sparse.csc_matrix(A),B).shape
Out[34]: (5, 3)

with A,B dense matrices:
In [42]: A.shape
Out[42]: (3, 5)
In [43]: B.shape
Out[43]: (5, 3)
In [44]: dot(A,B).shape
Out[44]: (3, 3)
In [45]: dot(sparse.csc_matrix(A),sparse.csc_matrix(B)).shape
Out[45]: (3, 3)
In [46]: dot(A,sparse.csc_matrix(B)).shape
Out[46]: (3, 5)
In [47]: dot(sparse.csc_matrix(A),B).shape
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)

/Users/robince/<ipython console> in <module>()

AttributeError: 'NotImplementedType' object has no attribute 'shape'

Cheers

Robin


From wnbell at gmail.com  Wed Oct  1 12:29:59 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Wed, 1 Oct 2008 12:29:59 -0400
Subject: [SciPy-user] dot behaviour with sparse matrix
In-Reply-To: <b14b627d0810010920m4a994a1q23feb756397c66c3@mail.gmail.com>
References: <b14b627d0810010920m4a994a1q23feb756397c66c3@mail.gmail.com>
Message-ID: <d05265cb0810010929x1e86be68h41bac962fe98fa5e@mail.gmail.com>

On Wed, Oct 1, 2008 at 12:20 PM, Robin <robince at gmail.com> wrote:
> Not sure if this is a bug or not but it was certainly unexpected
> behaviour for me.
>
> I found that dot(dense,dense) works as expected
> (ie dimensions (8,255),(255,8) -> (8,8))
> and that dot(sparse,sparse) also works that way, producing a sparse result.
>
> But dot(dense,sparse) gives an output with the same dimensions as the
> dense input. Not sure why this is happening and I found it a bit
> confusing. Is this a bug or inteded behaviour?
>
> Actually I was just going through an example and I notice that
> dot(csc_matrix, dense_matrix) raises a NotImplemented type - so
> perhaps the other cases should also:
>

The sparse formats do support sparse * sparse, sparse * dense, and
dense * sparse using the standard infix operator.  I'm surprised that
dot() works at all with sparse arguments.

If you can get to the bottom of this I'd be happy to fix it.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From robert.kern at gmail.com  Wed Oct  1 15:39:04 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 1 Oct 2008 14:39:04 -0500
Subject: [SciPy-user] r4763 won't compile, missing lsodar.pyf
In-Reply-To: <2b1c8c4f0810010559v6defd4bbsd510798cfd47a55c@mail.gmail.com>
References: <2b1c8c4f0810010559v6defd4bbsd510798cfd47a55c@mail.gmail.com>
Message-ID: <3d375d730810011239l10cbeb1cr2f123685b1e1e304@mail.gmail.com>

On Wed, Oct 1, 2008 at 07:59, James Philbin <philbinj at gmail.com> wrote:
> Hi,
>
> "python setup.py build"
> errors with:
> "target build/src.linux-x86_64-2.5/lsodarmodule.c does not exist:
>   Assuming lsodarmodule.c was generated with "build_src --inplace" command."
>
> And indeed, scipy/integrate/setup.py has:
> "config.add_extension('lsodar',
>                         sources=['lsodar.pyf'],
>                         libraries=libs,
>                         **newblas)"
>
> But scipy/integrate doesn't contain lsodar.pyf?

It was a mistaken commit. Fixed in r4764.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From wesmckinn at gmail.com  Thu Oct  2 14:18:45 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Thu, 2 Oct 2008 14:18:45 -0400
Subject: [SciPy-user] Bug in scipy.stats skew, kurtosis
Message-ID: <6c476c8a0810021118w654cf45s50194b22e7c1f2cb@mail.gmail.com>

Right now scipy.stats.skew (normal) and scipy.stats.kurtosis (fisher =
False) applied to a 1-dimensional array return an ndarray of the single
value with no dimensions.

In [13]: scipy.stats.skew([1,2,3,4,5])
Out[13]: array(0.0)

I am using Scipy 0.6, but I looked at the latest SVN and this behavior has
not changed. It's being caused by these lines:

m2 = moment(a, 2, axis)
m3 = moment(a, 3, axis)
zero = (m2 == 0)
vals = np.where(zero, 0, m3 / m2**1.5)

when the condition is not an array in numpy.where, the result is this 0-dim
array. Not sure if this is a SciPy issue or a numpy issue. If it's a Scipy
issue I can put in a ticket for this. I am using a workaround for now (check
if the result of the calc is an array and has no shape, if so cast to float)
but this seems like a bug.

Thanks,
Wes
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081002/79207183/attachment.html>

From josef.pktd at gmail.com  Thu Oct  2 16:37:20 2008
From: josef.pktd at gmail.com (joep)
Date: Thu, 2 Oct 2008 13:37:20 -0700 (PDT)
Subject: [SciPy-user] Bug in scipy.stats skew, kurtosis
In-Reply-To: <6c476c8a0810021118w654cf45s50194b22e7c1f2cb@mail.gmail.com>
References: <6c476c8a0810021118w654cf45s50194b22e7c1f2cb@mail.gmail.com>
Message-ID: <3d5c2fc1-c1f8-4514-8680-962344e258cc@25g2000hsk.googlegroups.com>



On Oct 2, 2:18?pm, "Wes McKinney" <wesmck... at gmail.com> wrote:
> Right now scipy.stats.skew (normal) and scipy.stats.kurtosis (fisher =
> False) applied to a 1-dimensional array return an ndarray of the single
> value with no dimensions.
>
> In [13]: scipy.stats.skew([1,2,3,4,5])
> Out[13]: array(0.0)
>
> I am using Scipy 0.6, but I looked at the latest SVN and this behavior has
> not changed. It's being caused by these lines:
>
> m2 = moment(a, 2, axis)
> m3 = moment(a, 3, axis)
> zero = (m2 == 0)
> vals = np.where(zero, 0, m3 / m2**1.5)
>
> when the condition is not an array in numpy.where, the result is this 0-dim
> array. Not sure if this is a SciPy issue or a numpy issue. If it's a Scipy
> issue I can put in a ticket for this. I am using a workaround for now (check
> if the result of the calc is an array and has no shape, if so cast to float)
> but this seems like a bug.
>
> Thanks,
> Wes
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-u... at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user


In Changeset 4752, this has been changed for pdf, cdf, and so on,
however other methods such as stats still have the same property:

>>> stats.norm.cdf(0.4)
0.65542174161032418

>>> stats.norm.stats(moments='s')
array(0.0)
>>> stats.norm.stats(moments='s')[()]
0.0
>>> stats.norm.stats(moments='s').shape
()

For consistency, it might be better to apply this change to all
methods in scipy.stats that return 0-dim array.

Josef


From nwagner at iam.uni-stuttgart.de  Fri Oct  3 02:35:16 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 03 Oct 2008 08:35:16 +0200
Subject: [SciPy-user] Python Optimization Modeling Objects (Pyomo)
Message-ID: <web-112244124@uni-stuttgart.de>

FWIW,
  
http://www.optimization-online.org/DB_HTML/2008/09/2095.html

Nils



From dmitrey.kroshko at scipy.org  Fri Oct  3 03:15:21 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Fri, 03 Oct 2008 10:15:21 +0300
Subject: [SciPy-user] Python Optimization Modeling Objects (Pyomo)
In-Reply-To: <web-112244124@uni-stuttgart.de>
References: <web-112244124@uni-stuttgart.de>
Message-ID: <48E5C689.5020508@scipy.org>

Thank you Nils, I was wondering why my daily visitors number this week 
jumped from ~ 50 to ~ 80-90, and now I know the answer: in the article 
they mentioned
"However, Coopr Opt is not as mature as the OpenOpt package"
:)

Unfortunately, some days ago I committed a nasty bug and now I'm hunting 
for the one because my latest svn doesn't work properly with NL problems.

D.

Nils Wagner wrote:
> FWIW,
>   
> http://www.optimization-online.org/DB_HTML/2008/09/2095.html
>
> Nils
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From nwagner at iam.uni-stuttgart.de  Fri Oct  3 04:40:59 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 03 Oct 2008 10:40:59 +0200
Subject: [SciPy-user] Python Optimization Modeling Objects (Pyomo)
In-Reply-To: <48E5C689.5020508@scipy.org>
References: <web-112244124@uni-stuttgart.de>
 <48E5C689.5020508@scipy.org>
Message-ID: <web-112249337@uni-stuttgart.de>

On Fri, 03 Oct 2008 10:15:21 +0300
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> Thank you Nils, I was wondering why my daily visitors 
>number this week 
> jumped from ~ 50 to ~ 80-90, and now I know the answer: 
>in the article 
> they mentioned
> "However, Coopr Opt is not as mature as the OpenOpt 
>package"
> :)
> 
> Unfortunately, some days ago I committed a nasty bug and 
>now I'm hunting 
> for the one because my latest svn doesn't work properly 
>with NL problems.
> 
> D.
> 
> Nils Wagner wrote:
>> FWIW,
>>   
>> http://www.optimization-online.org/DB_HTML/2008/09/2095.html
>>
>> Nils
>>
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>>
>>
>>   
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

Dmitrey,

You might be interested in

http://www.optimization-online.org/DB_HTML/2008/09/2101.html
(recent paper on pswarm)

as well.

Cheers,
             Nils
  


From jeremy at jeremysanders.net  Fri Oct  3 04:49:43 2008
From: jeremy at jeremysanders.net (Jeremy Sanders)
Date: Fri, 03 Oct 2008 09:49:43 +0100
Subject: [SciPy-user] ANN: Veusz 1.1
Message-ID: <gc4mb8$o3m$1@ger.gmane.org>

Veusz 1.1
---------
Velvet Ember Under Sky Zenith
-----------------------------
http://home.gna.org/veusz/

Veusz is Copyright (C) 2003-2008 Jeremy Sanders <jeremy at jeremysanders.net>
Licenced under the GPL (version 2 or greater).

Veusz is a scientific plotting package written in Python, using PyQt4
for display and user-interfaces, and numpy for handling the numeric
data. Veusz is designed to produce publication-ready Postscript/PDF
output. The user interface aims to be simple, consistent and powerful.

Veusz provides a GUI, command line, embedding and scripting interface
(based on Python) to its plotting facilities. It also allows for
manipulation and editing of datasets.

Feature changes from 1.0:
 * Axes autoscale when plotting functions
 * Labels can be dragged around on plots
 * More marker symbols
 * SVG export of plots

 * The point plotting and axis range code has been rewritten.
 * Includes quite a few minor bugfixes
 
Features of package:
 * X-Y plots (with errorbars)
 * Line and function plots
 * Contour plots
 * Images (with colour mappings and colorbars)
 * Stepped plots (for histograms)
 * Fitting functions to data
 * Stacked plots and arrays of plots
 * Plot keys
 * Plot labels
 * LaTeX-like formatting for text
 * EPS/PDF/PNG export
 * Scripting interface
 * Dataset creation/manipulation
 * Embed Veusz within other programs
 * Text, CSV and FITS importing

Requirements:
 Python (2.3 or greater required)
   http://www.python.org/
 Qt >= 4.3 (free edition)
   http://www.trolltech.com/products/qt/  
 PyQt >= 4.3 (SIP is required to be installed first)
   http://www.riverbankcomputing.co.uk/pyqt/
   http://www.riverbankcomputing.co.uk/sip/
 numpy >= 1.0
   http://numpy.scipy.org/
 Microsoft Core Fonts (recommended for nice output)
   http://corefonts.sourceforge.net/
 PyFITS >= 1.1 (optional for FITS import)
   http://www.stsci.edu/resources/software_hardware/pyfits

For documentation on using Veusz, see the "Documents" directory. The
manual is in pdf, html and text format (generated from docbook).

Issues:
 * Can be very slow to plot large datasets if antialiasing is enabled.
   Right click on graph and disable antialias to speed up output.
 * The embedding interface appears to crash on exiting.

If you enjoy using Veusz, I would love to hear from you. Please join
the mailing lists at

https://gna.org/mail/?group=veusz

to discuss new features or if you'd like to contribute code. The
latest code can always be found in the SVN repository.

Jeremy Sanders




From jrs65 at cam.ac.uk  Fri Oct  3 13:06:23 2008
From: jrs65 at cam.ac.uk (Richard Shaw)
Date: Fri, 03 Oct 2008 18:06:23 +0100
Subject: [SciPy-user] Python 2.6 compilation error
Message-ID: <48E6510F.8060808@cam.ac.uk>

Hello,

I'm trying to compile Scipy 0.6 for Python 2.6 and I'm suffering from a 
compilation error in
scipy/sparse/linalg/dsolve/SuperLU/SRC/scomplex.h with conflicting types 
for ?_Py_c_abs?. In fact the same error as this bug:

http://scipy.org/scipy/scipy/ticket/735

Is there a workaround I could use in the meantime, rather than 
downgrading to Python 2.5? Anything would be appreciated.

Thanks,
Richard


From dhsu2 at u.washington.edu  Fri Oct  3 17:27:53 2008
From: dhsu2 at u.washington.edu (David Hsu)
Date: Fri, 3 Oct 2008 21:27:53 +0000 (UTC)
Subject: [SciPy-user]
	=?utf-8?q?Installation_from_source_on_OS_X=3A_=27Non?=
	=?utf-8?q?eType=27=09object_has_no_attribute_=27link=5Fshared=5Fob?=
	=?utf-8?q?ject=27?=
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
	<3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
	<BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>
	<3d375d730809231153p43290a4bu10cdf119908c8830@mail.gmail.com>
	<20080926225534.GA6402@astro.ox.ac.uk>
	<3d375d730809261614j7696ced5je18889036ea67fa1@mail.gmail.com>
	<5B3D66A5-B141-4DE7-99B3-A53F5EEE00E8@astro.ox.ac.uk>
Message-ID: <loom.20081003T212736-113@post.gmane.org>

Michael Williams <williams <at> astro.ox.ac.uk> writes:

> 
> 
> On 27 Sep 2008, at 00:14, Robert Kern wrote:
> > I very much recommend against using the binaries from HPC. They
> > release binaries for buggy bleeding-edge versions of gfortran, and
> > don't keep previous versions around.
> 
> Thanks for the heads-up. I'm not experienced with Fortran.
> 
> > I have had *much* more luck with
> > the binaries over here:
> >
> >  http://r.research.att.com/tools/
> >
> > Since you don't have admin privileges, you need to do a little command
> > line work instead of just being able to use the installer. Mount
> > gfortran-4.2.3.dmg. At the terminal:
> 
> That procedure to install the compiler worked fine, and scipy itself  
> seems to have built cleanly using it.
> 
> Thanks very much!
> 


Hi, I'm having the exact same problem and error message, but for some reason I
can't get past this step.  I've tried the suggestions here:

http://www.scipy.org/Installing_SciPy/Mac_OS_X

At the step where it requires building the fftw library, I get:

DRIVE:fftw-3.1.2 dhsu$ make
make  all-recursive
Making all in support
make[2]: *** No rule to make target `all'.  Stop.
make[1]: *** [all-recursive] Error 1
make: *** [all] Error 2

and "sudo make install" has the same result.  Then, when I go to build SciPy
itself, I get almost the exact same errors as in the thread above.  What am I
doing wrong?

Thanks,

David


Specifications:

MacBook Pro Core Duo
Python 2.5.1
Numpy 1.0.4
SciPy 0.6.0
GNU Fortran 4.2.3 (as suggested)




From oldcanine at yahoo.com  Sat Oct  4 07:56:04 2008
From: oldcanine at yahoo.com (Barry Olddog)
Date: Sat, 4 Oct 2008 04:56:04 -0700 (PDT)
Subject: [SciPy-user] undefined symbols when trying to import packages with
	linalg
Message-ID: <127988.64002.qm@web59605.mail.ac4.yahoo.com>

I've been struggling with building scipy into python2.5 for a couple
days on a new Centos 5.2, 64-bit machine. I finally got everything
built. At first I installed the blas and lapack packages, and then
numpy, which seems ok, and finally scipy.  

The first problem
was getting scipy to find the blas and lapack libraries, and then
various build errors. I removed the packages, and built Atlas myself
with lapack. Now scipy builds, and importing just scipy alone is ok,
but it balks at importing some of the packages, including stats,
linalg, optimize. Here's the complete import error:

>>> import scipy.stats
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python2.5/site-packages/scipy/stats/__init__.py", line 7,
in <module>
    from stats import *
  File "/usr/local/lib/python2.5/site-packages/scipy/stats/stats.py", line 192, 
in <module>
    import scipy.linalg as linalg
  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/__init__.py", line 8
, in <module>
    from basic import *
  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/basic.py", line 17, 
in <module>
    from lapack import get_lapack_funcs
  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/lapack.py", line 17,
in <module>
    from scipy.linalg import flapack
ImportError: /usr/local/lib/python2.5/site-packages/scipy/linalg/flapack.so: und
efined symbol: _gfortran_st_write_done

I
had thought that I was using g77 consistently. Earlier, I was able to
find references to conflicts betwen g77 and gfortran, which is why I
built my own atlas, etc. So the _gfortran_st_write_done error seems odd. 

Any
suggestions?
I just did this on a similar machine without any problem whatsoever.
I'm not sure but  I maybe installed the atlas rpm on the earlier one.
Could that
have been what did it?

Thanks



      



From josef.pktd at gmail.com  Sat Oct  4 09:51:12 2008
From: josef.pktd at gmail.com (joep)
Date: Sat, 4 Oct 2008 06:51:12 -0700 (PDT)
Subject: [SciPy-user] call to vectorize works only after several calls
Message-ID: <d9310a99-32d9-4f71-bc0a-76780cec60da@l42g2000hsc.googlegroups.com>

I get some strange behavior, when vectorized function is called
without valid arguments:
The first time I call it, it throws an exception, if I call it again,
after I called a with a non-empyty argument list, then it works. I
don't see why the underlying state should have changed.

Any clues?
This makes testing and debugging pretty difficult

Josef

------------- file: test_ppf_indexerror.py ----------------------
from scipy import stats

print '\nfirst try'
try:
    print stats.foldcauchy.ppf(0,0.2)
except Exception,e:
    print 'error occured on first try'
    print e
print '\ncalling with vec arguments'
print stats.foldcauchy.ppf([-1,0.0, 0.1,0.9,1.0,2],0.2)
print '\nsecond try'
print stats.foldcauchy.ppf(0,0.2)
print ' second try worked'
------------------------------------------------

when called from command line, this produces (with trunk and scipy
0.6.0, numpy 1.2rc2):

{{{
>python test_ppf_indexerror.py

first try
error occured on first try
invalid index

calling with vec arguments
[        NaN  0.          0.16455889  6.31992596         Inf
NaN]

second try
0.0
 second try worked
}}}

Here is the traceback if I call it just once on the command line:

{{{
python -c "from scipy import stats;print stats.foldcauchy.ppf(0,0.2)"
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "C:\Josef\_progs\virtualpy25\envscipy\lib\site-packages\scipy
\stats\distr
ibutions.py", line 587, in ppf
    place(output,cond,self._ppf(*goodargs)*scale + loc)
  File "C:\Josef\_progs\virtualpy25\envscipy\lib\site-packages\scipy
\stats\distr
ibutions.py", line 399, in _ppf
    return self.vecfunc(q,*args)
  File "C:\Programs\Python25\Lib\site-packages\numpy\lib
\function_base.py", line
 1648, in __call__
    newargs.append(asarray(arg).flat[0])
IndexError: invalid index
}}}


From wnbell at gmail.com  Sat Oct  4 18:39:08 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Sat, 4 Oct 2008 18:39:08 -0400
Subject: [SciPy-user] Python 2.6 compilation error
In-Reply-To: <48E6510F.8060808@cam.ac.uk>
References: <48E6510F.8060808@cam.ac.uk>
Message-ID: <d05265cb0810041539h6e5a1efgfb9b81960ecd19af@mail.gmail.com>

On Fri, Oct 3, 2008 at 1:06 PM, Richard Shaw <jrs65 at cam.ac.uk> wrote:
>
> I'm trying to compile Scipy 0.6 for Python 2.6 and I'm suffering from a
> compilation error in
> scipy/sparse/linalg/dsolve/SuperLU/SRC/scomplex.h with conflicting types
> for '_Py_c_abs'. In fact the same error as this bug:
>
> http://scipy.org/scipy/scipy/ticket/735
>
> Is there a workaround I could use in the meantime, rather than
> downgrading to Python 2.5? Anything would be appreciated.
>

Hi Richard, thanks for the report.  I've committed a workaround to
scipy r4767 in SVN so this issue doesn't appear in the coming SciPy
0.7 release.  I'd ask you to try using this version to see if the bug
is indeed fixed.  If you'd like to apply the same workaround to 0.6
you can use the following script:

sparse/linalg/dsolve/SuperLU/SRC$ for f in *; do sed
's/c_abs/slu_c_abs/g' < $f > temp; mv temp $f; done;

It simply renames c_abs() and c_abs1() to slu_c_abs() and slu_c_abs1()
respectively.

We really need to update our SuperLU code to a more recent release.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From robert.kern at gmail.com  Sat Oct  4 19:34:38 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Sat, 4 Oct 2008 18:34:38 -0500
Subject: [SciPy-user] undefined symbols when trying to import packages
	with linalg
In-Reply-To: <127988.64002.qm@web59605.mail.ac4.yahoo.com>
References: <127988.64002.qm@web59605.mail.ac4.yahoo.com>
Message-ID: <3d375d730810041634o282334dao7fe8dac3a972ad8@mail.gmail.com>

On Sat, Oct 4, 2008 at 06:56, Barry Olddog <oldcanine at yahoo.com> wrote:
> I've been struggling with building scipy into python2.5 for a couple
> days on a new Centos 5.2, 64-bit machine. I finally got everything
> built. At first I installed the blas and lapack packages, and then
> numpy, which seems ok, and finally scipy.
>
> The first problem
> was getting scipy to find the blas and lapack libraries, and then
> various build errors. I removed the packages, and built Atlas myself
> with lapack. Now scipy builds, and importing just scipy alone is ok,
> but it balks at importing some of the packages, including stats,
> linalg, optimize. Here's the complete import error:
>
>>>> import scipy.stats
> Traceback (most recent call last):
>  File "<stdin>", line 1, in <module>
>  File "/usr/local/lib/python2.5/site-packages/scipy/stats/__init__.py", line 7,
> in <module>
>    from stats import *
>  File "/usr/local/lib/python2.5/site-packages/scipy/stats/stats.py", line 192,
> in <module>
>    import scipy.linalg as linalg
>  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/__init__.py", line 8
> , in <module>
>    from basic import *
>  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/basic.py", line 17,
> in <module>
>    from lapack import get_lapack_funcs
>  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/lapack.py", line 17,
> in <module>
>    from scipy.linalg import flapack
> ImportError: /usr/local/lib/python2.5/site-packages/scipy/linalg/flapack.so: und
> efined symbol: _gfortran_st_write_done
>
> I
> had thought that I was using g77 consistently. Earlier, I was able to
> find references to conflicts betwen g77 and gfortran, which is why I
> built my own atlas, etc. So the _gfortran_st_write_done error seems odd.
>
> Any
> suggestions?

Use ldd on scipy/linalg/flapack.so to see what shared libraries it is
trying to link with. It should also show you exactly which .so files
it manages to find for the given library. You will probably see that
scipy/linalg/flapack.so is looking for libgfortran but not finding it.
If not, then keep using ldd on the found shared libraries until you
find the culprit. That is the library that was accidentally built with
gfortran.

If it's scipy/linalg/flapack.so, then look over your build log again
(and rebuild if necessary). Look near the beginning where
numpy.distutils is telling you what Fortran compilers it is finding.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From cournape at gmail.com  Sun Oct  5 01:59:08 2008
From: cournape at gmail.com (David Cournapeau)
Date: Sun, 5 Oct 2008 14:59:08 +0900
Subject: [SciPy-user] Installation from source on OS X: 'NoneType'
	object has no attribute 'link_shared_object'
In-Reply-To: <loom.20081003T212736-113@post.gmane.org>
References: <4B321E75-C644-48DD-96ED-71AEF9FE9CB3@astro.ox.ac.uk>
	<3d375d730809231039t239263e1wcf67da11c650c9b9@mail.gmail.com>
	<BBAAA7E7-7976-4FBD-A200-E1ED05ED70FF@astro.ox.ac.uk>
	<3d375d730809231153p43290a4bu10cdf119908c8830@mail.gmail.com>
	<20080926225534.GA6402@astro.ox.ac.uk>
	<3d375d730809261614j7696ced5je18889036ea67fa1@mail.gmail.com>
	<5B3D66A5-B141-4DE7-99B3-A53F5EEE00E8@astro.ox.ac.uk>
	<loom.20081003T212736-113@post.gmane.org>
Message-ID: <5b8d13220810042259g69c87cf9s28a48fcea18a0597@mail.gmail.com>

On Sat, Oct 4, 2008 at 6:27 AM, David Hsu <dhsu2 at u.washington.edu> wrote:
>
> Hi, I'm having the exact same problem and error message, but for some reason I
> can't get past this step.  I've tried the suggestions here:
>
> http://www.scipy.org/Installing_SciPy/Mac_OS_X

It does not really answer your question, but fftw is an optional
dependency for scipy. You don't need to install it to have a working
scipy install.

cheers,

David


From nwagner at iam.uni-stuttgart.de  Sun Oct  5 05:29:52 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sun, 05 Oct 2008 11:29:52 +0200
Subject: [SciPy-user] Installing nose on Windows XP
Message-ID: <web-112316216@uni-stuttgart.de>

Hi all,
  
How do I install nose on Windows ?
This is my first attempt to run numpy/scipy on Windows.
The problem is that I cannot run numpy.test() on Windows.

Nils


From david at ar.media.kyoto-u.ac.jp  Sun Oct  5 05:19:44 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Sun, 05 Oct 2008 18:19:44 +0900
Subject: [SciPy-user] Installing nose on Windows XP
In-Reply-To: <web-112316216@uni-stuttgart.de>
References: <web-112316216@uni-stuttgart.de>
Message-ID: <48E886B0.4040302@ar.media.kyoto-u.ac.jp>

Nils Wagner wrote:
> Hi all,
>   
> How do I install nose on Windows ?
> This is my first attempt to run numpy/scipy on Windows.
> The problem is that I cannot run numpy.test() on Windows.

You can use easy_install:

easy_install nose

You have to use the -U if you want to upgrade to a new version if you
already have it installed:

easy_install -U nose

You can also simply install it from sources (it is pure python, with no
dependency, so just a matter of python setup.py install)

http://somethingaboutorange.com/mrl/projects/nose/



From fredmfp at gmail.com  Sun Oct  5 08:37:17 2008
From: fredmfp at gmail.com (fred)
Date: Sun, 05 Oct 2008 14:37:17 +0200
Subject: [SciPy-user] [numpy distutils] cpu detected...
Message-ID: <48E8B4FD.10503@gmail.com>

Hi,

Looking at numpy/distutils/cpuinfo.py (numpy 1.1.0)
I wonder why Core2 is only detected for 64 bits arch, and not for 32 bits.

I have a Core2 and fortran codes run faster when -xT flag is specified 
for 32 bits arch.

Any hint ?

TIA.


Cheers,

-- 
Fred


From fredmfp at gmail.com  Sun Oct  5 09:25:57 2008
From: fredmfp at gmail.com (fred)
Date: Sun, 05 Oct 2008 15:25:57 +0200
Subject: [SciPy-user] [numpy distutils] cpu detected...
In-Reply-To: <48E8B4FD.10503@gmail.com>
References: <48E8B4FD.10503@gmail.com>
Message-ID: <48E8C065.7070703@gmail.com>

fred a ?crit :
> Hi,
> 
> Looking at numpy/distutils/cpuinfo.py (numpy 1.1.0)
> I wonder why Core2 is only detected for 64 bits arch, and not for 32 bits.
> 
> I have a Core2 and fortran codes run faster when -xT flag is specified 
> for 32 bits arch.
> 
> Any hint ?
In fact, I really don't understand.

Knowing that IntelEM64TCompiler class inherit from IntelFCompiler in
numpy/distutils/fcompiler/intel.py,
get_flags_arch() redefines (line 165) IntelFCompiler's method.
So my Core2 can't be detected on my 64 bits arch and thus
IntelEM64TFCompiler's get_flags_arch() returns an empty list.

Do I misunderstand something trivial ???


Cheers,

-- 
Fred


From oldcanine at yahoo.com  Sun Oct  5 11:19:25 2008
From: oldcanine at yahoo.com (Barry Olddog)
Date: Sun, 5 Oct 2008 08:19:25 -0700 (PDT)
Subject: [SciPy-user] undefined symbols when trying to import packages
	with linalg
Message-ID: <664056.52116.qm@web59611.mail.ac4.yahoo.com>

----- Original Message ----

> From: Robert Kern <robert.kern at gmail.com>
> To: SciPy Users List <scipy-user at scipy.org>
> Sent: Saturday, October 4, 2008 7:34:38 PM
> Subject: Re: [SciPy-user] undefined symbols when trying to import packages with linalg
> 
> On Sat, Oct 4, 2008 at 06:56, Barry Olddog wrote:
> > I've been struggling with building scipy into python2.5 for a couple
> > days on a new Centos 5.2, 64-bit machine. I finally got everything
> > built. At first I installed the blas and lapack packages, and then
> > numpy, which seems ok, and finally scipy.
> >
> > The first problem
> > was getting scipy to find the blas and lapack libraries, and then
> > various build errors. I removed the packages, and built Atlas myself
> > with lapack. Now scipy builds, and importing just scipy alone is ok,
> > but it balks at importing some of the packages, including stats,
> > linalg, optimize. Here's the complete import error:
> >
> >>>> import scipy.stats
> > Traceback (most recent call last):
> >  File "", line 1, in 
> >  File "/usr/local/lib/python2.5/site-packages/scipy/stats/__init__.py", line 
> 7,
> > in 
> >    from stats import *
> >  File "/usr/local/lib/python2.5/site-packages/scipy/stats/stats.py", line 192,
> > in 
> >    import scipy.linalg as linalg
> >  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/__init__.py", line 
> 8
> > , in 
> >    from basic import *
> >  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/basic.py", line 17,
> > in 
> >    from lapack import get_lapack_funcs
> >  File "/usr/local/lib/python2.5/site-packages/scipy/linalg/lapack.py", line 
> 17,
> > in 
> >    from scipy.linalg import flapack
> > ImportError: /usr/local/lib/python2.5/site-packages/scipy/linalg/flapack.so: 
> und
> > efined symbol: _gfortran_st_write_done
> >
> > I
> > had thought that I was using g77 consistently. Earlier, I was able to
> > find references to conflicts betwen g77 and gfortran, which is why I
> > built my own atlas, etc. So the _gfortran_st_write_done error seems odd.
> >
> > Any
> > suggestions?
> 
> Use ldd on scipy/linalg/flapack.so to see what shared libraries it is
> trying to link with. It should also show you exactly which .so files
> it manages to find for the given library. You will probably see that
> scipy/linalg/flapack.so is looking for libgfortran but not finding it.
> If not, then keep using ldd on the found shared libraries until you
> find the culprit. That is the library that was accidentally built with
> gfortran.
> 
> If it's scipy/linalg/flapack.so, then look over your build log again
> (and rebuild if necessary). Look near the beginning where
> numpy.distutils is telling you what Fortran compilers it is finding.
> 
> -- 
> Robert Kern
> 

Thanks for the reply. It pointed me to a way out. I tried to track down the reference with ldd but never did. But I had compiled the full lapack library as a static library, and my only guess is that somewhere in there was the offending symbol. The config in lapack seemed to me to be clear about g77, but I took a drastic measure and solved the problem.

It may be inelegant, but I moved gfortran out of /usr/bin, leaving g77 as the only choice, and then rebuilt everything -- lapack with blas, atlas and finally scipy

Barry


      



From jrs65 at cam.ac.uk  Sun Oct  5 11:28:37 2008
From: jrs65 at cam.ac.uk (Richard Shaw)
Date: Sun, 05 Oct 2008 16:28:37 +0100
Subject: [SciPy-user] Python 2.6 compilation error
In-Reply-To: <d05265cb0810041539h6e5a1efgfb9b81960ecd19af@mail.gmail.com>
References: <48E6510F.8060808@cam.ac.uk>
	<d05265cb0810041539h6e5a1efgfb9b81960ecd19af@mail.gmail.com>
Message-ID: <48E8DD25.5060400@cam.ac.uk>

Nathan Bell wrote:

> Hi Richard, thanks for the report.  I've committed a workaround to
> scipy r4767 in SVN so this issue doesn't appear in the coming SciPy
> 0.7 release. 

Thanks Nathan that works a treat, it's all compiled and installed.

I do get an error when running scipy.test(), though it seems to be wioth 
the testing code, scipy itself seems to run fine. I've attached the 
message and traceback below.

Thanks again,
Richard

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test_error.txt
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081005/de683fd8/attachment.txt>

From xavier.gnata at gmail.com  Sun Oct  5 15:01:55 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Sun, 5 Oct 2008 21:01:55 +0200
Subject: [SciPy-user] Install scipy on ubnutu 8.10
Message-ID: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>

Hi,

I have installed an ubuntu 8.10 within a kvm image.
The goal is to write an as simple as possible procedure to compile
scipy on this distribution.
I would like to be able to compile scipy only after having installed
some packages. No tricks.
Good news, g77 is not needed anymore :)

I have one issue:
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to
find 'umfpack.h'
I have installed libsuitesparse-dev providing /usr/include/suitesparse/umfpack.h
Of course, I can solve this with a symlink but it is very ugly.
It looks like there is something to fix in scipy auto-detection
because this "usr/include/suitesparse/umfpack.h" should be detected
and used.

Once again, my goal is to end up with something like:
1) apt-get install "the correct list of packages"
2) compile/install scipy svn *wthout any extra config*.
3) you have a nice scipy installed and you can even  check that using
scipy.test() (ok ok it is an svn version so if you get some errors
please report them ;))

Xavier


From wnbell at gmail.com  Sun Oct  5 15:35:19 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Sun, 5 Oct 2008 15:35:19 -0400
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
Message-ID: <d05265cb0810051235i3a60d000p23f7ae9f7f35d0e4@mail.gmail.com>

On Sun, Oct 5, 2008 at 3:01 PM, Xavier Gnata <xavier.gnata at gmail.com> wrote:
>
> I have installed an ubuntu 8.10 within a kvm image.
> The goal is to write an as simple as possible procedure to compile
> scipy on this distribution.
> I would like to be able to compile scipy only after having installed
> some packages. No tricks.

Hi Xavier, I think this is a great idea.  I have sold other people on
numpy/scipy, but myriad BLAS/LAPACK/fortran combinations makes the
installation process unpleasant for new users.

> I have one issue:
> scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to
> find 'umfpack.h'
> I have installed libsuitesparse-dev providing /usr/include/suitesparse/umfpack.h
> Of course, I can solve this with a symlink but it is very ugly.
> It looks like there is something to fix in scipy auto-detection
> because this "usr/include/suitesparse/umfpack.h" should be detected
> and used.

The "right" way to do this is to put a [umfpack] section in site.cfg
http://projects.scipy.org/pipermail/scipy-dev/2007-April/006910.html

Ideally, auto-detection would always include knowledge about where the
major distributions put these files.  Keep in mind that scipy's
support for UMFPACK is deprecated in 0.7 and has been moved to a
scikit.

>
> Once again, my goal is to end up with something like:
> 1) apt-get install "the correct list of packages"
> 2) compile/install scipy svn *wthout any extra config*.
> 3) you have a nice scipy installed and you can even  check that using
> scipy.test() (ok ok it is an svn version so if you get some errors
> please report them ;))
>

For 2) you can either improve the auto-detection (I don't know how
myself) or simply not use UMFPACK.  The danger is that someone might
already have libsuitesparse-dev installed, so fixing the
auto-detection is probably a better approach.

Anyway, I'm eager to see this information become available.  Do you
plan to provide instructions for both 32-bit and 64-bit versions?  In
8.04 I believe there were some differences in the availability of BLAS
libraries between the two.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From xavier.gnata at gmail.com  Sun Oct  5 16:15:35 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Sun, 05 Oct 2008 22:15:35 +0200
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <d05265cb0810051235i3a60d000p23f7ae9f7f35d0e4@mail.gmail.com>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
	<d05265cb0810051235i3a60d000p23f7ae9f7f35d0e4@mail.gmail.com>
Message-ID: <48E92067.3080908@gmail.com>

Nathan Bell wrote:
> On Sun, Oct 5, 2008 at 3:01 PM, Xavier Gnata <xavier.gnata at gmail.com> wrote:
>   
>> I have installed an ubuntu 8.10 within a kvm image.
>> The goal is to write an as simple as possible procedure to compile
>> scipy on this distribution.
>> I would like to be able to compile scipy only after having installed
>> some packages. No tricks.
>>     
>
> Hi Xavier, I think this is a great idea.  I have sold other people on
> numpy/scipy, but myriad BLAS/LAPACK/fortran combinations makes the
> installation process unpleasant for new users.
>
>   
same here :(

>> I have one issue:
>> scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to
>> find 'umfpack.h'
>> I have installed libsuitesparse-dev providing /usr/include/suitesparse/umfpack.h
>> Of course, I can solve this with a symlink but it is very ugly.
>> It looks like there is something to fix in scipy auto-detection
>> because this "usr/include/suitesparse/umfpack.h" should be detected
>> and used.
>>     
>
> The "right" way to do this is to put a [umfpack] section in site.cfg
> http://projects.scipy.org/pipermail/scipy-dev/2007-April/006910.html
>
> Ideally, auto-detection would always include knowledge about where the
> major distributions put these files.  Keep in mind that scipy's
> support for UMFPACK is deprecated in 0.7 and has been moved to a
> scikit.
>
>   
Well I would like to have the feedback of the developers on the best  
way to fix the auto-detection.

site.cfg should not be needed here.


>> Once again, my goal is to end up with something like:
>> 1) apt-get install "the correct list of packages"
>> 2) compile/install scipy svn *wthout any extra config*.
>> 3) you have a nice scipy installed and you can even  check that using
>> scipy.test() (ok ok it is an svn version so if you get some errors
>> please report them ;))
>>
>>     
>
> For 2) you can either improve the auto-detection (I don't know how
> myself) or simply not use UMFPACK.  The danger is that someone might
> already have libsuitesparse-dev installed, so fixing the
> auto-detection is probably a better approach.
>
> Anyway, I'm eager to see this information become available.  Do you
> plan to provide instructions for both 32-bit and 64-bit versions?  In
> 8.04 I believe there were some differences in the availability of BLAS
> libraries between the two.
>
>   
I do not need umfpack but my goal is to perform a scipy installation 
with all the libs users could need.

I should be the same for 32 and 64 using 8.10 but I have to check that 
(with kvm it should be easy)
My plan is to provide an installation procedure *without an hand 
compiled ATLAS*.

Ubuntu is going to release 8.10 end of october. It would be nice to have 
a standard way to compile the svn on this "intrepid ibex".
Of course, the simplest  way to  having something working is  apt-get 
install python-scipy. The drawback with these nice packages it that you 
cannot ask for patch or test a new feature.

If someone wants to write a procedure to compile a full optimized scipy 
using  ATLAS  on 8.10 , he's really welcome :)

Xavier


From wnbell at gmail.com  Sun Oct  5 16:22:59 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Sun, 5 Oct 2008 16:22:59 -0400
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <48E92067.3080908@gmail.com>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
	<d05265cb0810051235i3a60d000p23f7ae9f7f35d0e4@mail.gmail.com>
	<48E92067.3080908@gmail.com>
Message-ID: <d05265cb0810051322wbd72fber74ce501c43f49ff4@mail.gmail.com>

On Sun, Oct 5, 2008 at 4:15 PM, Xavier Gnata <xavier.gnata at gmail.com> wrote:
>
> I should be the same for 32 and 64 using 8.10 but I have to check that
> (with kvm it should be easy)
> My plan is to provide an installation procedure *without an hand
> compiled ATLAS*.
>

The prepackaged ATLAS libraries should be sufficient for most users.
My understanding is that these support SSE2, which is the most
important SIMD advancement.

> Ubuntu is going to release 8.10 end of october. It would be nice to have
> a standard way to compile the svn on this "intrepid ibex".
> Of course, the simplest  way to  having something working is  apt-get
> install python-scipy. The drawback with these nice packages it that you
> cannot ask for patch or test a new feature.

Yep, we should have both.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From robert.kern at gmail.com  Sun Oct  5 17:08:41 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Sun, 5 Oct 2008 16:08:41 -0500
Subject: [SciPy-user] Python 2.6 compilation error
In-Reply-To: <48E8DD25.5060400@cam.ac.uk>
References: <48E6510F.8060808@cam.ac.uk>
	<d05265cb0810041539h6e5a1efgfb9b81960ecd19af@mail.gmail.com>
	<48E8DD25.5060400@cam.ac.uk>
Message-ID: <3d375d730810051408q2b7e428cq3c1e3aba2c91c337@mail.gmail.com>

On Sun, Oct 5, 2008 at 10:28, Richard Shaw <jrs65 at cam.ac.uk> wrote:
> Nathan Bell wrote:
>
>> Hi Richard, thanks for the report.  I've committed a workaround to
>> scipy r4767 in SVN so this issue doesn't appear in the coming SciPy
>> 0.7 release.
>
> Thanks Nathan that works a treat, it's all compiled and installed.
>
> I do get an error when running scipy.test(), though it seems to be wioth the
> testing code, scipy itself seems to run fine. I've attached the message and
> traceback below.

nose 0.10.3 isn't compatible with Python 2.6. Upgrade to 0.10.4.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From bblais at bryant.edu  Sun Oct  5 21:46:25 2008
From: bblais at bryant.edu (Brian Blais)
Date: Sun, 5 Oct 2008 21:46:25 -0400
Subject: [SciPy-user] where to post this tool?
Message-ID: <2D72079F-9E8A-4DE1-AF85-A23581B5E97A@bryant.edu>

Hello,

I am working on a tool which is a wrapper around odeint, but allows a  
more convenient input of the equations, and was wondering where would  
be a good place to post such a thing (when it is a bit more stable/ 
mature)?  Right now, I can do things like:

     sim=Simulation()
     sim.add("p'=a*p*(1-p/K)",100,plot=True)
     sim.params(a=1.5,K=300)

     sim.run(0,50)

for logistic growth, but also things like the lorenz model:

s=Simulation()

# lorenz model
s.add("C' = sig*(L-C)",13,plot=1)
s.add("L' = r*C-L-C*M",8.1)
s.add("M' = C*L-b*M",45,plot=2)
s.params(r=28.0,sig=10.0,b=8.0/3.0)

s.run(0,30)

I can do some higher order equations, and I hope to be able to do  
vector equations in the same way.

Anyway, I think there might be some who would be interested in this  
tool, and was wondering if there is a repository for such things.  It  
isn't so big to be a scikit, but too large to be a simple example in  
a cookbook I think, but the scipy cookbook looks like it might be  
close to the right place.  in matlab, there was a matlab repository  
for user-contributed m-files.  is there such a thing for python/scipy/ 
numpy?



			thanks,

				Brian Blais

-- 
Brian Blais
bblais at bryant.edu
http://web.bryant.edu/~bblais



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081005/6d425c54/attachment.html>

From david at ar.media.kyoto-u.ac.jp  Mon Oct  6 08:28:42 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 06 Oct 2008 21:28:42 +0900
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
Message-ID: <48EA047A.2060504@ar.media.kyoto-u.ac.jp>

Xavier Gnata wrote:
> Hi,
>
> I have installed an ubuntu 8.10 within a kvm image.
> The goal is to write an as simple as possible procedure to compile
> scipy on this distribution.
> I would like to be able to compile scipy only after having installed
> some packages. No tricks.
> Good news, g77 is not needed anymore :)

It was already the case for 8.04, but it was a bit confusing: both g77
and gfortran libraries were available. Is it better with 8.10 ?

>
> I have one issue:
> scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to
> find 'umfpack.h'
> I have installed libsuitesparse-dev providing /usr/include/suitesparse/umfpack.h
> Of course, I can solve this with a symlink but it is very ugly.
> It looks like there is something to fix in scipy auto-detection
> because this "usr/include/suitesparse/umfpack.h" should be detected
> and used.

That's really the debian packagers' fault. Why do they think it is a
good idea to change the header path of the library is beyond me; it
breaks every single package which depends on it. That's stupid.

We could get around it; but I though umfpack was being deprecated in
scipy  (that is, we would do a scikit from it, but scipy would not
depend on it anymore).

>
> Once again, my goal is to end up with something like:
> 1) apt-get install "the correct list of packages"
> 2) compile/install scipy svn *wthout any extra config*.
> 3) you have a nice scipy installed and you can even  check that using
> scipy.test() (ok ok it is an svn version so if you get some errors
> please report them ;))

It has been the case for a long time :) On old ubuntu:

sudo apt-get install g77 gcc python-dev atlas3-base-dev

On more recent ones:

sudo apt-get install gcc gfortran python-dev libatlas3gf-sse2-dev

The real solution would be to prived our own deb, though.

cheers,

David


From cdcasey at gmail.com  Mon Oct  6 12:19:13 2008
From: cdcasey at gmail.com (chris)
Date: Mon, 6 Oct 2008 11:19:13 -0500
Subject: [SciPy-user] Some failing tests
Message-ID: <ead057790810060919j76fbc0ebqf1e65d71a8903a5b@mail.gmail.com>

when i run scipy.test() on RedHat 32-bit 3 & 4, I get the following
feedback. Are these things I need to worry about? It's running scipy
0.6.0. Not sure if these are fixed in the next version, or if there's
a way I can fix them now....


>>> import scipy
>>> scipy.test()
Failed importing scipy.linsolve.umfpack: 'module' object has no
attribute 'umfpack'
  Found 9/9 tests for scipy.cluster.tests.test_vq
  Found 20/20 tests for scipy.fftpack.tests.test_pseudo_diffs
  Found 4/4 tests for scipy.fftpack.tests.test_helper
  Found 18/18 tests for scipy.fftpack.tests.test_basic
  Found 3/3 tests for scipy.integrate.tests.test_quadrature
  Found 1/1 tests for scipy.integrate.tests.test_integrate
  Found 10/10 tests for scipy.integrate.tests.test_quadpack
  Found 6/6 tests for scipy.tests.test_interpolate
  Found 6/6 tests for scipy.tests.test_fitpack
  Found 13/13 tests for scipy.io.tests.test_mmio
  Found 4/4 tests for scipy.io.tests.test_recaster
  Found 4/4 tests for scipy.io.tests.test_array_import
  Found 5/5 tests for scipy.io.tests.test_npfile
  Found 28/28 tests for scipy.io.tests.test_mio
  Found 128/128 tests for scipy.lib.blas.tests.test_fblas
  Found 16/16 tests for scipy.lib.blas.tests.test_blas

****************************************************************
WARNING: clapack module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses flapack instead of clapack.
****************************************************************

  Found 42/42 tests for scipy.lib.lapack.tests.test_lapack
  Found 6/6 tests for scipy.linalg.tests.test_iterative
  Found 41/41 tests for scipy.linalg.tests.test_basic
  Found 128/128 tests for scipy.linalg.tests.test_fblas
  Found 72/72 tests for scipy.linalg.tests.test_decomp
  Found 4/4 tests for scipy.linalg.tests.test_lapack
<module 'scipy.linalg.fblas' from
'/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linalg/fblas.so'>
  Found 16/16 tests for scipy.linalg.tests.test_blas
  Found 7/7 tests for scipy.linalg.tests.test_matfuncs
Failed importing
/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linsolve/umfpack/tests/test_umfpack.py:
'module' object has no attribute 'umfpack'
  Found 2/2 tests for scipy.maxentropy.tests.test_maxentropy
  Found 3/3 tests for scipy.tests.test_pilutil
  Found 395/395 tests for scipy.ndimage.tests.test_ndimage
  Found 5/5 tests for scipy.odr.tests.test_odr
  Found 8/8 tests for scipy.optimize.tests.test_optimize
  Found 1/1 tests for scipy.optimize.tests.test_cobyla
  Found 10/10 tests for scipy.optimize.tests.test_nonlin
  Found 4/4 tests for scipy.optimize.tests.test_zeros
  Found 5/5 tests for scipy.signal.tests.test_signaltools
  Found 4/4 tests for scipy.signal.tests.test_wavelets
  Found 152/152 tests for scipy.sparse.tests.test_sparse
  Found 342/342 tests for scipy.special.tests.test_basic
  Found 3/3 tests for scipy.special.tests.test_spfun_stats
  Found 10/10 tests for scipy.stats.tests.test_morestats
  Found 107/107 tests for scipy.stats.tests.test_stats
  Found 73/73 tests for scipy.stats.tests.test_distributions
building extensions here: /home/ccasey/.python25_compiled/m10
  Found 1/1 tests for scipy.weave.tests.test_ext_tools
  Found 0/0 tests for scipy.weave.tests.test_c_spec
  Found 74/74 tests for scipy.weave.tests.test_size_check
  Found 2/2 tests for scipy.weave.tests.test_blitz_tools
  Found 0/0 tests for scipy.weave.tests.test_scxx_object
  Found 0/0 tests for scipy.weave.tests.test_scxx_dict
  Found 0/0 tests for scipy.weave.tests.test_scxx_sequence
  Found 26/26 tests for scipy.weave.tests.test_catalog
  Found 1/1 tests for scipy.weave.tests.test_ast_tools
  Found 0/0 tests for scipy.weave.tests.test_inline_tools
  Found 3/3 tests for scipy.weave.tests.test_standard_array_spec
Failed importing
/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/weave/tests/test_wx_spec.py:
Could not locate wxPython base directory.
  Found 16/16 tests for scipy.weave.tests.test_slice_handler
  Found 9/9 tests for scipy.weave.tests.test_build_tools
.../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/cluster/vq.py:477:
UserWarning: One of the clusters is empty. Re-run kmean with a
different initialization.
  warnings.warn("One of the clusters is empty. "
exception raised as expected: One of the clusters is empty. Re-run
kmean with a different initialization.
...................................................Residual: 1.05006950608e-07
................./home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/interpolate/fitpack2.py:458:
UserWarning:
The coefficients of the spline returned have been computed as the
minimal norm least-squares solution of a (numerically) rank deficient
system (deficiency=7). If deficiency is large, the results may be
inaccurate. Deficiency may strongly depend on the value of eps.
  warnings.warn(message)
.......................
Don't worry about a warning regarding the number of bytes read.
Warning: 1000000 bytes requested, 20 bytes read.
........................................caxpy:n=4
..caxpy:n=3
....ccopy:n=4
..ccopy:n=3
.............cscal:n=4
....cswap:n=4
..cswap:n=3
.....daxpy:n=4
..daxpy:n=3
....dcopy:n=4
..dcopy:n=3
.............dscal:n=4
....dswap:n=4
..dswap:n=3
.....saxpy:n=4
..saxpy:n=3
....scopy:n=4
..scopy:n=3
.............sscal:n=4
....sswap:n=4
..sswap:n=3
.....zaxpy:n=4
..zaxpy:n=3
....zcopy:n=4
..zcopy:n=3
.............zscal:n=4
....zswap:n=4
..zswap:n=3
............................................FF................................................................caxpy:n=4
..caxpy:n=3
....ccopy:n=4
..ccopy:n=3
.............cscal:n=4
....cswap:n=4
..cswap:n=3
.....daxpy:n=4
..daxpy:n=3
....dcopy:n=4
..dcopy:n=3
.............dscal:n=4
....dswap:n=4
..dswap:n=3
.....saxpy:n=4
..saxpy:n=3
....scopy:n=4
..scopy:n=3
.............sscal:n=4
....sswap:n=4
..sswap:n=3
.....zaxpy:n=4
..zaxpy:n=3
....zcopy:n=4
..zcopy:n=3
.............zscal:n=4
....zswap:n=4
..zswap:n=3
............................................................................
****************************************************************
WARNING: clapack module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses flapack instead of clapack.
****************************************************************

..
****************************************************************
WARNING: cblas module is empty
-----------
See scipy/INSTALL.txt for troubleshooting.
Notes:
* If atlas library is not found by numpy/distutils/system_info.py,
  then scipy uses fblas instead of cblas.
****************************************************************

.................Result may be inaccurate, approximate err = 5.71185593749e-09
................................................................................................................../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/ndimage/interpolation.py:41:
UserWarning: Mode "reflect" may yield incorrect results on boundaries.
Please use "mirror" instead.
  warnings.warn('Mode "reflect" may yield incorrect results on '
.................................................................................................................................................................................................................................................................................................................................................................Use
minimum degree ordering on A'+A.
.....................................Use minimum degree ordering on A'+A.
.....................................Use minimum degree ordering on A'+A.
................................Use minimum degree ordering on A'+A.
....................................................................................................................................................................................................................................................................................................................................................0.2
0.2
0.2
......0.2
..0.2
0.2
0.2
0.2
0.2
.....................Ties preclude use of exact statistic.
..Ties preclude use of exact statistic.
......................./home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:166:
FutureWarning:
        The semantics of histogram will be modified in
        release 1.2 to improve outlier handling. The new behavior can be
        obtained using new=True. Note that the new version accepts/returns
        the bin edges instead of the left bin edges.
        Please read the docstring for more information.
  Please read the docstring for more information.""", FutureWarning)
/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:181:
FutureWarning:
                Outliers handling will change in version 1.2.
                Please read the docstring for details.
  Please read the docstring for details.""", FutureWarning)
..............................................................................................................................................................................................................................................................................................warning:
specified build_dir '_bad_path_' does not exist or is not writable.
Trying default locations
...warning: specified build_dir '..' does not exist or is not
writable. Trying default locations
..warning: specified build_dir '_bad_path_' does not exist or is not
writable. Trying default locations
...warning: specified build_dir '..' does not exist or is not
writable. Trying default locations
.
======================================================================
FAIL: check_syevr (scipy.lib.lapack.tests.test_lapack.test_flapack_float)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
line 41, in check_syevr
    assert_array_almost_equal(w,exact_w)
  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
line 255, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
line 240, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])

======================================================================
FAIL: check_syevr_irange
(scipy.lib.lapack.tests.test_lapack.test_flapack_float)----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
line 66, in check_syevr_irange
    assert_array_almost_equal(w,exact_w[rslice])
  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
line 255, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
line 240, in assert_array_compare
    assert cond, msg
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])

----------------------------------------------------------------------
Ran 1847 tests in 6.389s

FAILED (failures=2)
<unittest._TextTestResult run=1847 errors=0 failures=2>


From cdcasey at gmail.com  Mon Oct  6 12:26:25 2008
From: cdcasey at gmail.com (chris)
Date: Mon, 6 Oct 2008 11:26:25 -0500
Subject: [SciPy-user] Some failing tests
In-Reply-To: <ead057790810060919j76fbc0ebqf1e65d71a8903a5b@mail.gmail.com>
References: <ead057790810060919j76fbc0ebqf1e65d71a8903a5b@mail.gmail.com>
Message-ID: <ead057790810060926t6e5d8basbf8f2ad9fae4f4be@mail.gmail.com>

Sorry. That should have read RedHat 4 32-bit (not 3 & 4). Also, using
numpy 1.1.1.

On Mon, Oct 6, 2008 at 11:19 AM, chris <cdcasey at gmail.com> wrote:
> when i run scipy.test() on RedHat 32-bit 3 & 4, I get the following
> feedback. Are these things I need to worry about? It's running scipy
> 0.6.0. Not sure if these are fixed in the next version, or if there's
> a way I can fix them now....
>
>
>>>> import scipy
>>>> scipy.test()
> Failed importing scipy.linsolve.umfpack: 'module' object has no
> attribute 'umfpack'
>  Found 9/9 tests for scipy.cluster.tests.test_vq
>  Found 20/20 tests for scipy.fftpack.tests.test_pseudo_diffs
>  Found 4/4 tests for scipy.fftpack.tests.test_helper
>  Found 18/18 tests for scipy.fftpack.tests.test_basic
>  Found 3/3 tests for scipy.integrate.tests.test_quadrature
>  Found 1/1 tests for scipy.integrate.tests.test_integrate
>  Found 10/10 tests for scipy.integrate.tests.test_quadpack
>  Found 6/6 tests for scipy.tests.test_interpolate
>  Found 6/6 tests for scipy.tests.test_fitpack
>  Found 13/13 tests for scipy.io.tests.test_mmio
>  Found 4/4 tests for scipy.io.tests.test_recaster
>  Found 4/4 tests for scipy.io.tests.test_array_import
>  Found 5/5 tests for scipy.io.tests.test_npfile
>  Found 28/28 tests for scipy.io.tests.test_mio
>  Found 128/128 tests for scipy.lib.blas.tests.test_fblas
>  Found 16/16 tests for scipy.lib.blas.tests.test_blas
>
> ****************************************************************
> WARNING: clapack module is empty
> -----------
> See scipy/INSTALL.txt for troubleshooting.
> Notes:
> * If atlas library is not found by numpy/distutils/system_info.py,
>  then scipy uses flapack instead of clapack.
> ****************************************************************
>
>  Found 42/42 tests for scipy.lib.lapack.tests.test_lapack
>  Found 6/6 tests for scipy.linalg.tests.test_iterative
>  Found 41/41 tests for scipy.linalg.tests.test_basic
>  Found 128/128 tests for scipy.linalg.tests.test_fblas
>  Found 72/72 tests for scipy.linalg.tests.test_decomp
>  Found 4/4 tests for scipy.linalg.tests.test_lapack
> <module 'scipy.linalg.fblas' from
> '/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linalg/fblas.so'>
>  Found 16/16 tests for scipy.linalg.tests.test_blas
>  Found 7/7 tests for scipy.linalg.tests.test_matfuncs
> Failed importing
> /home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linsolve/umfpack/tests/test_umfpack.py:
> 'module' object has no attribute 'umfpack'
>  Found 2/2 tests for scipy.maxentropy.tests.test_maxentropy
>  Found 3/3 tests for scipy.tests.test_pilutil
>  Found 395/395 tests for scipy.ndimage.tests.test_ndimage
>  Found 5/5 tests for scipy.odr.tests.test_odr
>  Found 8/8 tests for scipy.optimize.tests.test_optimize
>  Found 1/1 tests for scipy.optimize.tests.test_cobyla
>  Found 10/10 tests for scipy.optimize.tests.test_nonlin
>  Found 4/4 tests for scipy.optimize.tests.test_zeros
>  Found 5/5 tests for scipy.signal.tests.test_signaltools
>  Found 4/4 tests for scipy.signal.tests.test_wavelets
>  Found 152/152 tests for scipy.sparse.tests.test_sparse
>  Found 342/342 tests for scipy.special.tests.test_basic
>  Found 3/3 tests for scipy.special.tests.test_spfun_stats
>  Found 10/10 tests for scipy.stats.tests.test_morestats
>  Found 107/107 tests for scipy.stats.tests.test_stats
>  Found 73/73 tests for scipy.stats.tests.test_distributions
> building extensions here: /home/ccasey/.python25_compiled/m10
>  Found 1/1 tests for scipy.weave.tests.test_ext_tools
>  Found 0/0 tests for scipy.weave.tests.test_c_spec
>  Found 74/74 tests for scipy.weave.tests.test_size_check
>  Found 2/2 tests for scipy.weave.tests.test_blitz_tools
>  Found 0/0 tests for scipy.weave.tests.test_scxx_object
>  Found 0/0 tests for scipy.weave.tests.test_scxx_dict
>  Found 0/0 tests for scipy.weave.tests.test_scxx_sequence
>  Found 26/26 tests for scipy.weave.tests.test_catalog
>  Found 1/1 tests for scipy.weave.tests.test_ast_tools
>  Found 0/0 tests for scipy.weave.tests.test_inline_tools
>  Found 3/3 tests for scipy.weave.tests.test_standard_array_spec
> Failed importing
> /home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/weave/tests/test_wx_spec.py:
> Could not locate wxPython base directory.
>  Found 16/16 tests for scipy.weave.tests.test_slice_handler
>  Found 9/9 tests for scipy.weave.tests.test_build_tools
> .../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/cluster/vq.py:477:
> UserWarning: One of the clusters is empty. Re-run kmean with a
> different initialization.
>  warnings.warn("One of the clusters is empty. "
> exception raised as expected: One of the clusters is empty. Re-run
> kmean with a different initialization.
> ...................................................Residual: 1.05006950608e-07
> ................./home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/interpolate/fitpack2.py:458:
> UserWarning:
> The coefficients of the spline returned have been computed as the
> minimal norm least-squares solution of a (numerically) rank deficient
> system (deficiency=7). If deficiency is large, the results may be
> inaccurate. Deficiency may strongly depend on the value of eps.
>  warnings.warn(message)
> .......................
> Don't worry about a warning regarding the number of bytes read.
> Warning: 1000000 bytes requested, 20 bytes read.
> ........................................caxpy:n=4
> ..caxpy:n=3
> ....ccopy:n=4
> ..ccopy:n=3
> .............cscal:n=4
> ....cswap:n=4
> ..cswap:n=3
> .....daxpy:n=4
> ..daxpy:n=3
> ....dcopy:n=4
> ..dcopy:n=3
> .............dscal:n=4
> ....dswap:n=4
> ..dswap:n=3
> .....saxpy:n=4
> ..saxpy:n=3
> ....scopy:n=4
> ..scopy:n=3
> .............sscal:n=4
> ....sswap:n=4
> ..sswap:n=3
> .....zaxpy:n=4
> ..zaxpy:n=3
> ....zcopy:n=4
> ..zcopy:n=3
> .............zscal:n=4
> ....zswap:n=4
> ..zswap:n=3
> ............................................FF................................................................caxpy:n=4
> ..caxpy:n=3
> ....ccopy:n=4
> ..ccopy:n=3
> .............cscal:n=4
> ....cswap:n=4
> ..cswap:n=3
> .....daxpy:n=4
> ..daxpy:n=3
> ....dcopy:n=4
> ..dcopy:n=3
> .............dscal:n=4
> ....dswap:n=4
> ..dswap:n=3
> .....saxpy:n=4
> ..saxpy:n=3
> ....scopy:n=4
> ..scopy:n=3
> .............sscal:n=4
> ....sswap:n=4
> ..sswap:n=3
> .....zaxpy:n=4
> ..zaxpy:n=3
> ....zcopy:n=4
> ..zcopy:n=3
> .............zscal:n=4
> ....zswap:n=4
> ..zswap:n=3
> ............................................................................
> ****************************************************************
> WARNING: clapack module is empty
> -----------
> See scipy/INSTALL.txt for troubleshooting.
> Notes:
> * If atlas library is not found by numpy/distutils/system_info.py,
>  then scipy uses flapack instead of clapack.
> ****************************************************************
>
> ..
> ****************************************************************
> WARNING: cblas module is empty
> -----------
> See scipy/INSTALL.txt for troubleshooting.
> Notes:
> * If atlas library is not found by numpy/distutils/system_info.py,
>  then scipy uses fblas instead of cblas.
> ****************************************************************
>
> .................Result may be inaccurate, approximate err = 5.71185593749e-09
> ................................................................................................................../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/ndimage/interpolation.py:41:
> UserWarning: Mode "reflect" may yield incorrect results on boundaries.
> Please use "mirror" instead.
>  warnings.warn('Mode "reflect" may yield incorrect results on '
> .................................................................................................................................................................................................................................................................................................................................................................Use
> minimum degree ordering on A'+A.
> .....................................Use minimum degree ordering on A'+A.
> .....................................Use minimum degree ordering on A'+A.
> ................................Use minimum degree ordering on A'+A.
> ....................................................................................................................................................................................................................................................................................................................................................0.2
> 0.2
> 0.2
> ......0.2
> ..0.2
> 0.2
> 0.2
> 0.2
> 0.2
> .....................Ties preclude use of exact statistic.
> ..Ties preclude use of exact statistic.
> ......................./home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:166:
> FutureWarning:
>        The semantics of histogram will be modified in
>        release 1.2 to improve outlier handling. The new behavior can be
>        obtained using new=True. Note that the new version accepts/returns
>        the bin edges instead of the left bin edges.
>        Please read the docstring for more information.
>  Please read the docstring for more information.""", FutureWarning)
> /home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:181:
> FutureWarning:
>                Outliers handling will change in version 1.2.
>                Please read the docstring for details.
>  Please read the docstring for details.""", FutureWarning)
> ..............................................................................................................................................................................................................................................................................................warning:
> specified build_dir '_bad_path_' does not exist or is not writable.
> Trying default locations
> ...warning: specified build_dir '..' does not exist or is not
> writable. Trying default locations
> ..warning: specified build_dir '_bad_path_' does not exist or is not
> writable. Trying default locations
> ...warning: specified build_dir '..' does not exist or is not
> writable. Trying default locations
> .
> ======================================================================
> FAIL: check_syevr (scipy.lib.lapack.tests.test_lapack.test_flapack_float)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>  File "/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
> line 41, in check_syevr
>    assert_array_almost_equal(w,exact_w)
>  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
> line 255, in assert_array_almost_equal
>    header='Arrays are not almost equal')
>  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
> line 240, in assert_array_compare
>    assert cond, msg
> AssertionError:
> Arrays are not almost equal
>
> (mismatch 33.3333333333%)
>  x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
>  y: array([-0.66992434,  0.48769389,  9.18223045])
>
> ======================================================================
> FAIL: check_syevr_irange
> (scipy.lib.lapack.tests.test_lapack.test_flapack_float)----------------------------------------------------------------------
> Traceback (most recent call last):
>  File "/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
> line 66, in check_syevr_irange
>    assert_array_almost_equal(w,exact_w[rslice])
>  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
> line 255, in assert_array_almost_equal
>    header='Arrays are not almost equal')
>  File "/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
> line 240, in assert_array_compare
>    assert cond, msg
> AssertionError:
> Arrays are not almost equal
>
> (mismatch 33.3333333333%)
>  x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
>  y: array([-0.66992434,  0.48769389,  9.18223045])
>
> ----------------------------------------------------------------------
> Ran 1847 tests in 6.389s
>
> FAILED (failures=2)
> <unittest._TextTestResult run=1847 errors=0 failures=2>
>


From stefan at sun.ac.za  Mon Oct  6 12:28:21 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Mon, 6 Oct 2008 18:28:21 +0200
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
	<5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>
Message-ID: <9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>

2008/9/26 David Cournapeau <cournape at gmail.com>:
> Yes, it would be nice. What do other people think about deprecating
> all the numpy re-export in scipy ? It would be nice to do for 0.7
> (e.g. in 0.7, deprecated, in 0.8, removed).

There were no objections to this, so may we go ahead?

St?fan


From nwagner at iam.uni-stuttgart.de  Mon Oct  6 12:40:05 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Mon, 06 Oct 2008 18:40:05 +0200
Subject: [SciPy-user] Some failing tests
In-Reply-To: <ead057790810060926t6e5d8basbf8f2ad9fae4f4be@mail.gmail.com>
References: <ead057790810060919j76fbc0ebqf1e65d71a8903a5b@mail.gmail.com>
	<ead057790810060926t6e5d8basbf8f2ad9fae4f4be@mail.gmail.com>
Message-ID: <web-112387477@uni-stuttgart.de>

On Mon, 6 Oct 2008 11:26:25 -0500
  chris <cdcasey at gmail.com> wrote:
> Sorry. That should have read RedHat 4 32-bit (not 3 & 
>4). Also, using
> numpy 1.1.1.
> 
> On Mon, Oct 6, 2008 at 11:19 AM, chris 
><cdcasey at gmail.com> wrote:
>> when i run scipy.test() on RedHat 32-bit 3 & 4, I get 
>>the following
>> feedback. Are these things I need to worry about? It's 
>>running scipy
>> 0.6.0. Not sure if these are fixed in the next version, 
>>or if there's
>> a way I can fix them now....
>>
>>
>>>>> import scipy
>>>>> scipy.test()
>> Failed importing scipy.linsolve.umfpack: 'module' object 
>>has no
>> attribute 'umfpack'
>>  Found 9/9 tests for scipy.cluster.tests.test_vq
>>  Found 20/20 tests for 
>>scipy.fftpack.tests.test_pseudo_diffs
>>  Found 4/4 tests for scipy.fftpack.tests.test_helper
>>  Found 18/18 tests for scipy.fftpack.tests.test_basic
>>  Found 3/3 tests for 
>>scipy.integrate.tests.test_quadrature
>>  Found 1/1 tests for 
>>scipy.integrate.tests.test_integrate
>>  Found 10/10 tests for 
>>scipy.integrate.tests.test_quadpack
>>  Found 6/6 tests for scipy.tests.test_interpolate
>>  Found 6/6 tests for scipy.tests.test_fitpack
>>  Found 13/13 tests for scipy.io.tests.test_mmio
>>  Found 4/4 tests for scipy.io.tests.test_recaster
>>  Found 4/4 tests for scipy.io.tests.test_array_import
>>  Found 5/5 tests for scipy.io.tests.test_npfile
>>  Found 28/28 tests for scipy.io.tests.test_mio
>>  Found 128/128 tests for scipy.lib.blas.tests.test_fblas
>>  Found 16/16 tests for scipy.lib.blas.tests.test_blas
>>
>> ****************************************************************
>> WARNING: clapack module is empty
>> -----------
>> See scipy/INSTALL.txt for troubleshooting.
>> Notes:
>> * If atlas library is not found by 
>>numpy/distutils/system_info.py,
>>  then scipy uses flapack instead of clapack.
>> ****************************************************************
>>
>>  Found 42/42 tests for 
>>scipy.lib.lapack.tests.test_lapack
>>  Found 6/6 tests for scipy.linalg.tests.test_iterative
>>  Found 41/41 tests for scipy.linalg.tests.test_basic
>>  Found 128/128 tests for scipy.linalg.tests.test_fblas
>>  Found 72/72 tests for scipy.linalg.tests.test_decomp
>>  Found 4/4 tests for scipy.linalg.tests.test_lapack
>> <module 'scipy.linalg.fblas' from
>> '/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linalg/fblas.so'>
>>  Found 16/16 tests for scipy.linalg.tests.test_blas
>>  Found 7/7 tests for scipy.linalg.tests.test_matfuncs
>> Failed importing
>> /home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linsolve/umfpack/tests/test_umfpack.py:
>> 'module' object has no attribute 'umfpack'
>>  Found 2/2 tests for 
>>scipy.maxentropy.tests.test_maxentropy
>>  Found 3/3 tests for scipy.tests.test_pilutil
>>  Found 395/395 tests for 
>>scipy.ndimage.tests.test_ndimage
>>  Found 5/5 tests for scipy.odr.tests.test_odr
>>  Found 8/8 tests for scipy.optimize.tests.test_optimize
>>  Found 1/1 tests for scipy.optimize.tests.test_cobyla
>>  Found 10/10 tests for scipy.optimize.tests.test_nonlin
>>  Found 4/4 tests for scipy.optimize.tests.test_zeros
>>  Found 5/5 tests for scipy.signal.tests.test_signaltools
>>  Found 4/4 tests for scipy.signal.tests.test_wavelets
>>  Found 152/152 tests for scipy.sparse.tests.test_sparse
>>  Found 342/342 tests for scipy.special.tests.test_basic
>>  Found 3/3 tests for 
>>scipy.special.tests.test_spfun_stats
>>  Found 10/10 tests for scipy.stats.tests.test_morestats
>>  Found 107/107 tests for scipy.stats.tests.test_stats
>>  Found 73/73 tests for 
>>scipy.stats.tests.test_distributions
>> building extensions here: 
>>/home/ccasey/.python25_compiled/m10
>>  Found 1/1 tests for scipy.weave.tests.test_ext_tools
>>  Found 0/0 tests for scipy.weave.tests.test_c_spec
>>  Found 74/74 tests for scipy.weave.tests.test_size_check
>>  Found 2/2 tests for scipy.weave.tests.test_blitz_tools
>>  Found 0/0 tests for scipy.weave.tests.test_scxx_object
>>  Found 0/0 tests for scipy.weave.tests.test_scxx_dict
>>  Found 0/0 tests for 
>>scipy.weave.tests.test_scxx_sequence
>>  Found 26/26 tests for scipy.weave.tests.test_catalog
>>  Found 1/1 tests for scipy.weave.tests.test_ast_tools
>>  Found 0/0 tests for scipy.weave.tests.test_inline_tools
>>  Found 3/3 tests for 
>>scipy.weave.tests.test_standard_array_spec
>> Failed importing
>> /home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/weave/tests/test_wx_spec.py:
>> Could not locate wxPython base directory.
>>  Found 16/16 tests for 
>>scipy.weave.tests.test_slice_handler
>>  Found 9/9 tests for scipy.weave.tests.test_build_tools
>> .../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/cluster/vq.py:477:
>> UserWarning: One of the clusters is empty. Re-run kmean 
>>with a
>> different initialization.
>>  warnings.warn("One of the clusters is empty. "
>> exception raised as expected: One of the clusters is 
>>empty. Re-run
>> kmean with a different initialization.
>> ...................................................Residual: 
>>1.05006950608e-07
>> ................./home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/interpolate/fitpack2.py:458:
>> UserWarning:
>> The coefficients of the spline returned have been 
>>computed as the
>> minimal norm least-squares solution of a (numerically) 
>>rank deficient
>> system (deficiency=7). If deficiency is large, the 
>>results may be
>> inaccurate. Deficiency may strongly depend on the value 
>>of eps.
>>  warnings.warn(message)
>> .......................
>> Don't worry about a warning regarding the number of 
>>bytes read.
>> Warning: 1000000 bytes requested, 20 bytes read.
>> ........................................caxpy:n=4
>> ..caxpy:n=3
>> ....ccopy:n=4
>> ..ccopy:n=3
>> .............cscal:n=4
>> ....cswap:n=4
>> ..cswap:n=3
>> .....daxpy:n=4
>> ..daxpy:n=3
>> ....dcopy:n=4
>> ..dcopy:n=3
>> .............dscal:n=4
>> ....dswap:n=4
>> ..dswap:n=3
>> .....saxpy:n=4
>> ..saxpy:n=3
>> ....scopy:n=4
>> ..scopy:n=3
>> .............sscal:n=4
>> ....sswap:n=4
>> ..sswap:n=3
>> .....zaxpy:n=4
>> ..zaxpy:n=3
>> ....zcopy:n=4
>> ..zcopy:n=3
>> .............zscal:n=4
>> ....zswap:n=4
>> ..zswap:n=3
>> ............................................FF................................................................caxpy:n=4
>> ..caxpy:n=3
>> ....ccopy:n=4
>> ..ccopy:n=3
>> .............cscal:n=4
>> ....cswap:n=4
>> ..cswap:n=3
>> .....daxpy:n=4
>> ..daxpy:n=3
>> ....dcopy:n=4
>> ..dcopy:n=3
>> .............dscal:n=4
>> ....dswap:n=4
>> ..dswap:n=3
>> .....saxpy:n=4
>> ..saxpy:n=3
>> ....scopy:n=4
>> ..scopy:n=3
>> .............sscal:n=4
>> ....sswap:n=4
>> ..sswap:n=3
>> .....zaxpy:n=4
>> ..zaxpy:n=3
>> ....zcopy:n=4
>> ..zcopy:n=3
>> .............zscal:n=4
>> ....zswap:n=4
>> ..zswap:n=3
>> ............................................................................
>> ****************************************************************
>> WARNING: clapack module is empty
>> -----------
>> See scipy/INSTALL.txt for troubleshooting.
>> Notes:
>> * If atlas library is not found by 
>>numpy/distutils/system_info.py,
>>  then scipy uses flapack instead of clapack.
>> ****************************************************************
>>
>> ..
>> ****************************************************************
>> WARNING: cblas module is empty
>> -----------
>> See scipy/INSTALL.txt for troubleshooting.
>> Notes:
>> * If atlas library is not found by 
>>numpy/distutils/system_info.py,
>>  then scipy uses fblas instead of cblas.
>> ****************************************************************
>>
>> .................Result may be inaccurate, approximate 
>>err = 5.71185593749e-09
>> ................................................................................................................../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/ndimage/interpolation.py:41:
>> UserWarning: Mode "reflect" may yield incorrect results 
>>on boundaries.
>> Please use "mirror" instead.
>>  warnings.warn('Mode "reflect" may yield incorrect 
>>results on '
>> .................................................................................................................................................................................................................................................................................................................................................................Use
>> minimum degree ordering on A'+A.
>> .....................................Use minimum degree 
>>ordering on A'+A.
>> .....................................Use minimum degree 
>>ordering on A'+A.
>> ................................Use minimum degree 
>>ordering on A'+A.
>> ....................................................................................................................................................................................................................................................................................................................................................0.2
>> 0.2
>> 0.2
>> ......0.2
>> ..0.2
>> 0.2
>> 0.2
>> 0.2
>> 0.2
>> .....................Ties preclude use of exact 
>>statistic.
>> ..Ties preclude use of exact statistic.
>> ......................./home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:166:
>> FutureWarning:
>>        The semantics of histogram will be modified in
>>        release 1.2 to improve outlier handling. The new 
>>behavior can be
>>        obtained using new=True. Note that the new 
>>version accepts/returns
>>        the bin edges instead of the left bin edges.
>>        Please read the docstring for more information.
>>  Please read the docstring for more information.""", 
>>FutureWarning)
>> /home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:181:
>> FutureWarning:
>>                Outliers handling will change in version 
>>1.2.
>>                Please read the docstring for details.
>>  Please read the docstring for details.""", 
>>FutureWarning)
>> ..............................................................................................................................................................................................................................................................................................warning:
>> specified build_dir '_bad_path_' does not exist or is 
>>not writable.
>> Trying default locations
>> ...warning: specified build_dir '..' does not exist or 
>>is not
>> writable. Trying default locations
>> ..warning: specified build_dir '_bad_path_' does not 
>>exist or is not
>> writable. Trying default locations
>> ...warning: specified build_dir '..' does not exist or 
>>is not
>> writable. Trying default locations
>> .
>> ======================================================================
>> FAIL: check_syevr 
>>(scipy.lib.lapack.tests.test_lapack.test_flapack_float)
>> ----------------------------------------------------------------------
>> Traceback (most recent call last):
>>  File 
>>"/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
>> line 41, in check_syevr
>>    assert_array_almost_equal(w,exact_w)
>>  File 
>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>> line 255, in assert_array_almost_equal
>>    header='Arrays are not almost equal')
>>  File 
>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>> line 240, in assert_array_compare
>>    assert cond, msg
>> AssertionError:
>> Arrays are not almost equal
>>
>> (mismatch 33.3333333333%)
>>  x: array([-0.66992444,  0.48769444,  9.18222618], 
>>dtype=float32)
>>  y: array([-0.66992434,  0.48769389,  9.18223045])
>>
>> ======================================================================
>> FAIL: check_syevr_irange
>> (scipy.lib.lapack.tests.test_lapack.test_flapack_float)----------------------------------------------------------------------
>> Traceback (most recent call last):
>>  File 
>>"/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
>> line 66, in check_syevr_irange
>>    assert_array_almost_equal(w,exact_w[rslice])
>>  File 
>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>> line 255, in assert_array_almost_equal
>>    header='Arrays are not almost equal')
>>  File 
>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>> line 240, in assert_array_compare
>>    assert cond, msg
>> AssertionError:
>> Arrays are not almost equal
>>
>> (mismatch 33.3333333333%)
>>  x: array([-0.66992444,  0.48769444,  9.18222618], 
>>dtype=float32)
>>  y: array([-0.66992434,  0.48769389,  9.18223045])
>>
>> ----------------------------------------------------------------------
>> Ran 1847 tests in 6.389s
>>
>> FAILED (failures=2)
>> <unittest._TextTestResult run=1847 errors=0 failures=2>
>>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user


This is a known issue
See
http://scipy.org/scipy/scipy/ticket/375

  
Nils


From cdcasey at gmail.com  Mon Oct  6 12:55:30 2008
From: cdcasey at gmail.com (chris)
Date: Mon, 6 Oct 2008 11:55:30 -0500
Subject: [SciPy-user] Some failing tests
In-Reply-To: <web-112387477@uni-stuttgart.de>
References: <ead057790810060919j76fbc0ebqf1e65d71a8903a5b@mail.gmail.com>
	<ead057790810060926t6e5d8basbf8f2ad9fae4f4be@mail.gmail.com>
	<web-112387477@uni-stuttgart.de>
Message-ID: <ead057790810060955j4d4b1200g40c0d230fcd17a40@mail.gmail.com>

Thanks!

On Mon, Oct 6, 2008 at 11:40 AM, Nils Wagner
<nwagner at iam.uni-stuttgart.de> wrote:
> On Mon, 6 Oct 2008 11:26:25 -0500
>  chris <cdcasey at gmail.com> wrote:
>> Sorry. That should have read RedHat 4 32-bit (not 3 &
>>4). Also, using
>> numpy 1.1.1.
>>
>> On Mon, Oct 6, 2008 at 11:19 AM, chris
>><cdcasey at gmail.com> wrote:
>>> when i run scipy.test() on RedHat 32-bit 3 & 4, I get
>>>the following
>>> feedback. Are these things I need to worry about? It's
>>>running scipy
>>> 0.6.0. Not sure if these are fixed in the next version,
>>>or if there's
>>> a way I can fix them now....
>>>
>>>
>>>>>> import scipy
>>>>>> scipy.test()
>>> Failed importing scipy.linsolve.umfpack: 'module' object
>>>has no
>>> attribute 'umfpack'
>>>  Found 9/9 tests for scipy.cluster.tests.test_vq
>>>  Found 20/20 tests for
>>>scipy.fftpack.tests.test_pseudo_diffs
>>>  Found 4/4 tests for scipy.fftpack.tests.test_helper
>>>  Found 18/18 tests for scipy.fftpack.tests.test_basic
>>>  Found 3/3 tests for
>>>scipy.integrate.tests.test_quadrature
>>>  Found 1/1 tests for
>>>scipy.integrate.tests.test_integrate
>>>  Found 10/10 tests for
>>>scipy.integrate.tests.test_quadpack
>>>  Found 6/6 tests for scipy.tests.test_interpolate
>>>  Found 6/6 tests for scipy.tests.test_fitpack
>>>  Found 13/13 tests for scipy.io.tests.test_mmio
>>>  Found 4/4 tests for scipy.io.tests.test_recaster
>>>  Found 4/4 tests for scipy.io.tests.test_array_import
>>>  Found 5/5 tests for scipy.io.tests.test_npfile
>>>  Found 28/28 tests for scipy.io.tests.test_mio
>>>  Found 128/128 tests for scipy.lib.blas.tests.test_fblas
>>>  Found 16/16 tests for scipy.lib.blas.tests.test_blas
>>>
>>> ****************************************************************
>>> WARNING: clapack module is empty
>>> -----------
>>> See scipy/INSTALL.txt for troubleshooting.
>>> Notes:
>>> * If atlas library is not found by
>>>numpy/distutils/system_info.py,
>>>  then scipy uses flapack instead of clapack.
>>> ****************************************************************
>>>
>>>  Found 42/42 tests for
>>>scipy.lib.lapack.tests.test_lapack
>>>  Found 6/6 tests for scipy.linalg.tests.test_iterative
>>>  Found 41/41 tests for scipy.linalg.tests.test_basic
>>>  Found 128/128 tests for scipy.linalg.tests.test_fblas
>>>  Found 72/72 tests for scipy.linalg.tests.test_decomp
>>>  Found 4/4 tests for scipy.linalg.tests.test_lapack
>>> <module 'scipy.linalg.fblas' from
>>> '/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linalg/fblas.so'>
>>>  Found 16/16 tests for scipy.linalg.tests.test_blas
>>>  Found 7/7 tests for scipy.linalg.tests.test_matfuncs
>>> Failed importing
>>> /home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/linsolve/umfpack/tests/test_umfpack.py:
>>> 'module' object has no attribute 'umfpack'
>>>  Found 2/2 tests for
>>>scipy.maxentropy.tests.test_maxentropy
>>>  Found 3/3 tests for scipy.tests.test_pilutil
>>>  Found 395/395 tests for
>>>scipy.ndimage.tests.test_ndimage
>>>  Found 5/5 tests for scipy.odr.tests.test_odr
>>>  Found 8/8 tests for scipy.optimize.tests.test_optimize
>>>  Found 1/1 tests for scipy.optimize.tests.test_cobyla
>>>  Found 10/10 tests for scipy.optimize.tests.test_nonlin
>>>  Found 4/4 tests for scipy.optimize.tests.test_zeros
>>>  Found 5/5 tests for scipy.signal.tests.test_signaltools
>>>  Found 4/4 tests for scipy.signal.tests.test_wavelets
>>>  Found 152/152 tests for scipy.sparse.tests.test_sparse
>>>  Found 342/342 tests for scipy.special.tests.test_basic
>>>  Found 3/3 tests for
>>>scipy.special.tests.test_spfun_stats
>>>  Found 10/10 tests for scipy.stats.tests.test_morestats
>>>  Found 107/107 tests for scipy.stats.tests.test_stats
>>>  Found 73/73 tests for
>>>scipy.stats.tests.test_distributions
>>> building extensions here:
>>>/home/ccasey/.python25_compiled/m10
>>>  Found 1/1 tests for scipy.weave.tests.test_ext_tools
>>>  Found 0/0 tests for scipy.weave.tests.test_c_spec
>>>  Found 74/74 tests for scipy.weave.tests.test_size_check
>>>  Found 2/2 tests for scipy.weave.tests.test_blitz_tools
>>>  Found 0/0 tests for scipy.weave.tests.test_scxx_object
>>>  Found 0/0 tests for scipy.weave.tests.test_scxx_dict
>>>  Found 0/0 tests for
>>>scipy.weave.tests.test_scxx_sequence
>>>  Found 26/26 tests for scipy.weave.tests.test_catalog
>>>  Found 1/1 tests for scipy.weave.tests.test_ast_tools
>>>  Found 0/0 tests for scipy.weave.tests.test_inline_tools
>>>  Found 3/3 tests for
>>>scipy.weave.tests.test_standard_array_spec
>>> Failed importing
>>> /home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/weave/tests/test_wx_spec.py:
>>> Could not locate wxPython base directory.
>>>  Found 16/16 tests for
>>>scipy.weave.tests.test_slice_handler
>>>  Found 9/9 tests for scipy.weave.tests.test_build_tools
>>> .../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/cluster/vq.py:477:
>>> UserWarning: One of the clusters is empty. Re-run kmean
>>>with a
>>> different initialization.
>>>  warnings.warn("One of the clusters is empty. "
>>> exception raised as expected: One of the clusters is
>>>empty. Re-run
>>> kmean with a different initialization.
>>> ...................................................Residual:
>>>1.05006950608e-07
>>> ................./home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/interpolate/fitpack2.py:458:
>>> UserWarning:
>>> The coefficients of the spline returned have been
>>>computed as the
>>> minimal norm least-squares solution of a (numerically)
>>>rank deficient
>>> system (deficiency=7). If deficiency is large, the
>>>results may be
>>> inaccurate. Deficiency may strongly depend on the value
>>>of eps.
>>>  warnings.warn(message)
>>> .......................
>>> Don't worry about a warning regarding the number of
>>>bytes read.
>>> Warning: 1000000 bytes requested, 20 bytes read.
>>> ........................................caxpy:n=4
>>> ..caxpy:n=3
>>> ....ccopy:n=4
>>> ..ccopy:n=3
>>> .............cscal:n=4
>>> ....cswap:n=4
>>> ..cswap:n=3
>>> .....daxpy:n=4
>>> ..daxpy:n=3
>>> ....dcopy:n=4
>>> ..dcopy:n=3
>>> .............dscal:n=4
>>> ....dswap:n=4
>>> ..dswap:n=3
>>> .....saxpy:n=4
>>> ..saxpy:n=3
>>> ....scopy:n=4
>>> ..scopy:n=3
>>> .............sscal:n=4
>>> ....sswap:n=4
>>> ..sswap:n=3
>>> .....zaxpy:n=4
>>> ..zaxpy:n=3
>>> ....zcopy:n=4
>>> ..zcopy:n=3
>>> .............zscal:n=4
>>> ....zswap:n=4
>>> ..zswap:n=3
>>> ............................................FF................................................................caxpy:n=4
>>> ..caxpy:n=3
>>> ....ccopy:n=4
>>> ..ccopy:n=3
>>> .............cscal:n=4
>>> ....cswap:n=4
>>> ..cswap:n=3
>>> .....daxpy:n=4
>>> ..daxpy:n=3
>>> ....dcopy:n=4
>>> ..dcopy:n=3
>>> .............dscal:n=4
>>> ....dswap:n=4
>>> ..dswap:n=3
>>> .....saxpy:n=4
>>> ..saxpy:n=3
>>> ....scopy:n=4
>>> ..scopy:n=3
>>> .............sscal:n=4
>>> ....sswap:n=4
>>> ..sswap:n=3
>>> .....zaxpy:n=4
>>> ..zaxpy:n=3
>>> ....zcopy:n=4
>>> ..zcopy:n=3
>>> .............zscal:n=4
>>> ....zswap:n=4
>>> ..zswap:n=3
>>> ............................................................................
>>> ****************************************************************
>>> WARNING: clapack module is empty
>>> -----------
>>> See scipy/INSTALL.txt for troubleshooting.
>>> Notes:
>>> * If atlas library is not found by
>>>numpy/distutils/system_info.py,
>>>  then scipy uses flapack instead of clapack.
>>> ****************************************************************
>>>
>>> ..
>>> ****************************************************************
>>> WARNING: cblas module is empty
>>> -----------
>>> See scipy/INSTALL.txt for troubleshooting.
>>> Notes:
>>> * If atlas library is not found by
>>>numpy/distutils/system_info.py,
>>>  then scipy uses fblas instead of cblas.
>>> ****************************************************************
>>>
>>> .................Result may be inaccurate, approximate
>>>err = 5.71185593749e-09
>>> ................................................................................................................../home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/ndimage/interpolation.py:41:
>>> UserWarning: Mode "reflect" may yield incorrect results
>>>on boundaries.
>>> Please use "mirror" instead.
>>>  warnings.warn('Mode "reflect" may yield incorrect
>>>results on '
>>> .................................................................................................................................................................................................................................................................................................................................................................Use
>>> minimum degree ordering on A'+A.
>>> .....................................Use minimum degree
>>>ordering on A'+A.
>>> .....................................Use minimum degree
>>>ordering on A'+A.
>>> ................................Use minimum degree
>>>ordering on A'+A.
>>> ....................................................................................................................................................................................................................................................................................................................................................0.2
>>> 0.2
>>> 0.2
>>> ......0.2
>>> ..0.2
>>> 0.2
>>> 0.2
>>> 0.2
>>> 0.2
>>> .....................Ties preclude use of exact
>>>statistic.
>>> ..Ties preclude use of exact statistic.
>>> ......................./home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:166:
>>> FutureWarning:
>>>        The semantics of histogram will be modified in
>>>        release 1.2 to improve outlier handling. The new
>>>behavior can be
>>>        obtained using new=True. Note that the new
>>>version accepts/returns
>>>        the bin edges instead of the left bin edges.
>>>        Please read the docstring for more information.
>>>  Please read the docstring for more information.""",
>>>FutureWarning)
>>> /home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/lib/function_base.py:181:
>>> FutureWarning:
>>>                Outliers handling will change in version
>>>1.2.
>>>                Please read the docstring for details.
>>>  Please read the docstring for details.""",
>>>FutureWarning)
>>> ..............................................................................................................................................................................................................................................................................................warning:
>>> specified build_dir '_bad_path_' does not exist or is
>>>not writable.
>>> Trying default locations
>>> ...warning: specified build_dir '..' does not exist or
>>>is not
>>> writable. Trying default locations
>>> ..warning: specified build_dir '_bad_path_' does not
>>>exist or is not
>>> writable. Trying default locations
>>> ...warning: specified build_dir '..' does not exist or
>>>is not
>>> writable. Trying default locations
>>> .
>>> ======================================================================
>>> FAIL: check_syevr
>>>(scipy.lib.lapack.tests.test_lapack.test_flapack_float)
>>> ----------------------------------------------------------------------
>>> Traceback (most recent call last):
>>>  File
>>>"/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
>>> line 41, in check_syevr
>>>    assert_array_almost_equal(w,exact_w)
>>>  File
>>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>>> line 255, in assert_array_almost_equal
>>>    header='Arrays are not almost equal')
>>>  File
>>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>>> line 240, in assert_array_compare
>>>    assert cond, msg
>>> AssertionError:
>>> Arrays are not almost equal
>>>
>>> (mismatch 33.3333333333%)
>>>  x: array([-0.66992444,  0.48769444,  9.18222618],
>>>dtype=float32)
>>>  y: array([-0.66992434,  0.48769389,  9.18223045])
>>>
>>> ======================================================================
>>> FAIL: check_syevr_irange
>>> (scipy.lib.lapack.tests.test_lapack.test_flapack_float)----------------------------------------------------------------------
>>> Traceback (most recent call last):
>>>  File
>>>"/home/ccasey/epd/lib/python2.5/site-packages/scipy-0.6.0.0006_s-py2.5-linux-i686.egg/scipy/lib/lapack/tests/esv_tests.py",
>>> line 66, in check_syevr_irange
>>>    assert_array_almost_equal(w,exact_w[rslice])
>>>  File
>>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>>> line 255, in assert_array_almost_equal
>>>    header='Arrays are not almost equal')
>>>  File
>>>"/home/ccasey/epd/lib/python2.5/site-packages/numpy-1.1.1.0001-py2.5-linux-i686.egg/numpy/testing/utils.py",
>>> line 240, in assert_array_compare
>>>    assert cond, msg
>>> AssertionError:
>>> Arrays are not almost equal
>>>
>>> (mismatch 33.3333333333%)
>>>  x: array([-0.66992444,  0.48769444,  9.18222618],
>>>dtype=float32)
>>>  y: array([-0.66992434,  0.48769389,  9.18223045])
>>>
>>> ----------------------------------------------------------------------
>>> Ran 1847 tests in 6.389s
>>>
>>> FAILED (failures=2)
>>> <unittest._TextTestResult run=1847 errors=0 failures=2>
>>>
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
> This is a known issue
> See
> http://scipy.org/scipy/scipy/ticket/375
>
>
> Nils
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From jah.mailinglist at gmail.com  Mon Oct  6 13:47:06 2008
From: jah.mailinglist at gmail.com (jah)
Date: Mon, 6 Oct 2008 10:47:06 -0700
Subject: [SciPy-user] Ubuntu Libraries
Message-ID: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>

Hi,

I was hoping someone could clarify the ubuntu packages and what is actually
needed by scipy.

1) I keep seeing posts about g77 not being required anymore and gfortran
preferred.  What is the "official" statement?  In INSTALL.txt, only g77 is
mentioned for Ubuntu packages via:  """Debian/Ubuntu packages (g77):
atlas3-base atlas3-base-dev""".  The only reference to gfortran is with Mac
OS X.  Also, in "Optional Packages" debian packages 'gcc g++ g77' are
recommened.

2) The install notes also say that a complete version of LAPACK is required
for scipy.  All the ubuntu descriptions say that only a subset of routines
from LAPACK are included with ATLAS.  For example,
http://packages.ubuntu.com/hardy/libatlas-sse2-dev.  Is this a problem?

3) It looks like there are two packages:  atlas3-sse2 and
libatlas-sse2-dev.  Are both required?  It seems that they refer to
different versions of ATLAS.

 http://packages.ubuntu.com/hardy/atlas3-sse2
 http://packages.ubuntu.com/hardy/libatlas-sse2-dev

On reading http://www.scipy.org/Installing_SciPy/Linux, it seems like I only
should need libatlas-sse2-dev.  Correct?

Thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081006/37aac4a3/attachment.html>

From peridot.faceted at gmail.com  Mon Oct  6 14:34:19 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Mon, 6 Oct 2008 14:34:19 -0400
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
	<5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>
	<9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>
Message-ID: <ce557a360810061134s1fe2cfa5t288284e5d8315b69@mail.gmail.com>

2008/10/6 St?fan van der Walt <stefan at sun.ac.za>:
> 2008/9/26 David Cournapeau <cournape at gmail.com>:
>> Yes, it would be nice. What do other people think about deprecating
>> all the numpy re-export in scipy ? It would be nice to do for 0.7
>> (e.g. in 0.7, deprecated, in 0.8, removed).
>
> There were no objections to this, so may we go ahead?

My only concern is possible user confusion: some functions (e.g. sqrt)
are provided as "enhanced" versions in scipy, while others are simply
reexported. If we remove the reexports, users can't simply use
scipy.whatever to get the best-available version of each function,
they have to know whether an enhanced version exists. Of course, since
the enhanced versions exist because their APIs differ in important and
possibly surprising ways (e.g., sqrt(-1) has a different return type
from sqrt(1)) this may be a good thing.

Anne

From haase at msg.ucsf.edu  Mon Oct  6 14:59:39 2008
From: haase at msg.ucsf.edu (Sebastian Haase)
Date: Mon, 6 Oct 2008 20:59:39 +0200
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <ce557a360810061134s1fe2cfa5t288284e5d8315b69@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
	<5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>
	<9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>
	<ce557a360810061134s1fe2cfa5t288284e5d8315b69@mail.gmail.com>
Message-ID: <bc657ead0810061159jd8d8b5cjf7f39a424c7a568d@mail.gmail.com>

On Mon, Oct 6, 2008 at 8:34 PM, Anne Archibald
<peridot.faceted at gmail.com> wrote:
> 2008/10/6 St?fan van der Walt <stefan at sun.ac.za>:
>> 2008/9/26 David Cournapeau <cournape at gmail.com>:
>>> Yes, it would be nice. What do other people think about deprecating
>>> all the numpy re-export in scipy ? It would be nice to do for 0.7
>>> (e.g. in 0.7, deprecated, in 0.8, removed).
>>
>> There were no objections to this, so may we go ahead?
>
> My only concern is possible user confusion: some functions (e.g. sqrt)
> are provided as "enhanced" versions in scipy, while others are simply
> reexported. If we remove the reexports, users can't simply use
> scipy.whatever to get the best-available version of each function,
> they have to know whether an enhanced version exists. Of course, since
> the enhanced versions exist because their APIs differ in important and
> possibly surprising ways (e.g., sqrt(-1) has a different return type
> from sqrt(1)) this may be a good thing.
>
Arguing that SciPy is below 1.0 I think the reexporting should be
minimized as much as possible.
I don't thing that some people's preference for
"from scipy import *"
(without a preceding "from numpy import *") should not be a deciding point.

It would be nice if it were clear enough (in general) if a given
function should be expected to part of numpy or part of scipy.
Left over uncertainties should get documented is a short form - a list maybe.
My two cents....
-Sebastian Haase


From rmay31 at gmail.com  Mon Oct  6 15:06:17 2008
From: rmay31 at gmail.com (Ryan May)
Date: Mon, 06 Oct 2008 14:06:17 -0500
Subject: [SciPy-user] Inconsistent standard deviation and variance
 implementation in scipy vs. scipy.stats
In-Reply-To: <9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>	<200809241205.02637.pgmdevlist@gmail.com>	<48DA9519.1000203@american.edu>	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>	<5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>
	<9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>
Message-ID: <48EA61A9.6040305@gmail.com>

St?fan van der Walt wrote:
> 2008/9/26 David Cournapeau <cournape at gmail.com>:
>> Yes, it would be nice. What do other people think about deprecating
>> all the numpy re-export in scipy ? It would be nice to do for 0.7
>> (e.g. in 0.7, deprecated, in 0.8, removed).
> 
> There were no objections to this, so may we go ahead?

+1

(My 0.02)

Ryan

-- 
Ryan May
Graduate Research Assistant
School of Meteorology
University of Oklahoma


From chiefmurph at hotmail.com  Mon Oct  6 16:23:42 2008
From: chiefmurph at hotmail.com (Dan Murphy)
Date: Mon, 6 Oct 2008 13:23:42 -0700
Subject: [SciPy-user] Gaussian quadrature error
In-Reply-To: <ce557a360809170900r4f2105b6g24089a81a6612150@mail.gmail.com>
References: <0F8253EA348F49F6A3C3A0AEFE044866@GatewayLaptop>
	<ce557a360809170900r4f2105b6g24089a81a6612150@mail.gmail.com>
Message-ID: <BLU149-W65E2810654A9A934CB2428DC390@phx.gbl>


Thanks for the clear explanation, Anne. If I understand the second take-home lesson -- feed integrators unity-order quantities -- then instead of calling quadrature as I did:
   integrate.quadrature(f,-1000.0,0.0)
I should scale my function f so that my call looks more like
   integrate.quadrature(f,-1.0,0.0)
Was that what you meant?
 
Also, when you say "scipy.optimize.quad is a little more general-purpose" did you mean "scipy.integrate.quad"? Indeed, the call
   integrate(quad(f,-1000.0,0.0)
worked great!
Thanks, and sorry for the late followup. 
 
Dan Murphy
> Date: Wed, 17 Sep 2008 12:00:09 -0400> From: peridot.faceted at gmail.com> To: scipy-user at scipy.org> Subject: Re: [SciPy-user] Gaussian quadrature error> > 2008/9/14 Dan Murphy <chiefmurph at comcast.net>:> > I am trying out the integrate.quadrature function on the function f(x)=e**x> > to the left of the y-axis. If the lower bound in not too negative, I get a> > reasonable answer, but if the lower bound is too negative, I get 0.0 as the> > value of the integral. Here is the code:> >> >> >> > from scipy import *> >> >> >> > def f(x):> >> > return e**x> >> >> >> > integrate.quadrature(f,-10.0,0.0) # answer is (0.999954600065,> > 3.14148596026e-010)> >> >> >> > but> >> >> >> > integrate.quadrature(f,-1000.0,0.0) # yields (8.35116510531e-090,> > 8.35116510531e-090)> >> >> >> > Note that 'val' and 'err' are equal. Is this a bug in quadrature?> > No, unfortunately. It is a limitation of numerical quadrature in> general. Specifically, no matter how adaptive the algorithm is, it can> only base its result on a finite number of sampled points of the> function. If these points are all zero to numerical accuracy, then the> answer must be zero. So if you imagine those samples are taken at the> midpoints of 10 intervals evenly spaced between -1000 and 0, then the> rightmost one returns a value of e**(-50), which is as close to zero> as makes no nevermind. You might be all right if this were an adaptive> scheme and if it used the endpoints, since one endpoint is guaranteed> to give you one. But not using the endpoints is a design feature of> some numerical integration schemes.> > The take-home lesson is that you can't just use numerical quadrature> systems blindly; you have to know the features and limitations of the> particular one you're using. Gaussian quadrature can be very accurate> for smooth functions, but it has a very specific domain of> applicability. scipy.optimize.quad is a little more general-purpose by> intent (and necessarily a little less efficient when Gaussian> quadrature will do) but it can be tricked too.> > A more specific take-home lesson is to try to normalize your problem> as much as possible, so that all quantities you feed your integrator> are of order unity. Yes, it's a pain to have to handle scale factors> yourself, particularly in the normal case when you're solving a family> of related problems. But you'll get much more reliable performance.> > Anne> _______________________________________________> SciPy-user mailing list> SciPy-user at scipy.org> http://projects.scipy.org/mailman/listinfo/scipy-user
_________________________________________________________________
Stay up to date on your PC, the Web, and your mobile phone with Windows Live.
http://clk.atdmt.com/MRT/go/msnnkwxp1020093185mrt/direct/01/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081006/c3f04b34/attachment.html>

From xavier.gnata at gmail.com  Mon Oct  6 18:59:15 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Tue, 07 Oct 2008 00:59:15 +0200
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <48EA047A.2060504@ar.media.kyoto-u.ac.jp>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
	<48EA047A.2060504@ar.media.kyoto-u.ac.jp>
Message-ID: <48EA9843.5000505@gmail.com>

>
>> Hi,
>>
>> I have installed an ubuntu 8.10 within a kvm image.
>> The goal is to write an as simple as possible procedure to compile
>> scipy on this distribution.
>> I would like to be able to compile scipy only after having installed
>> some packages. No tricks.
>> Good news, g77 is not needed anymore :)
>>     
>
> It was already the case for 8.04, but it was a bit confusing: both g77
> and gfortran libraries were available. Is it better with 8.10 ?
>
>   

looks like it is better.

>> I have one issue:
>> scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to
>> find 'umfpack.h'
>> I have installed libsuitesparse-dev providing /usr/include/suitesparse/umfpack.h
>> Of course, I can solve this with a symlink but it is very ugly.
>> It looks like there is something to fix in scipy auto-detection
>> because this "usr/include/suitesparse/umfpack.h" should be detected
>> and used.
>>     
>
> That's really the debian packagers' fault. Why do they think it is a
> good idea to change the header path of the library is beyond me; it
> breaks every single package which depends on it. That's stupid.
>
>   
ok. I'm going to write a bug report. w&s.

> We could get around it; but I though umfpack was being deprecated in
> scipy  (that is, we would do a scikit from it, but scipy would not
> depend on it anymore).
>
>   
Well it is good news :) but even At revision 4786, I get this:

umfpack_info:
  libraries umfpack not found in /usr/lib
/usr/lib/python2.5/site-packages/numpy/distutils/system_info.py:414: 
UserWarning:
    UMFPACK sparse solver (http://www.cise.ufl.edu/research/sparse/umfpack/)
    not found. Directories to search for the libraries can be specified 
in the
    numpy/distutils/site.cfg file (section [umfpack]) or by setting
    the UMFPACK environment variable.
  warnings.warn(self.notfounderror.__doc__)
  NOT AVAILABLE

on anorther 8.04 (and not 8.10) ubuntu.

>> Once again, my goal is to end up with something like:
>> 1) apt-get install "the correct list of packages"
>> 2) compile/install scipy svn *wthout any extra config*.
>> 3) you have a nice scipy installed and you can even  check that using
>> scipy.test() (ok ok it is an svn version so if you get some errors
>> please report them ;))
>>     
>
> It has been the case for a long time :) On old ubuntu:
>
> sudo apt-get install g77 gcc python-dev atlas3-base-dev
>
> On more recent ones:
>
> sudo apt-get install gcc gfortran python-dev libatlas3gf-sse2-dev
>
>   
Except if you have installed libsuiteparse- 3.1.0
You also need g++ don't you?
On my cucrrent install of 8.10, I cannot install libatlas3gf-sse2 but 
only libatlas3gf-base but it is only a beta.

scipy is a pretty uncommon piece of software, it is a scientific one :)
All the guys using scipy I know are using the svn version because they 
*need* *this* so nice feature.
Quite often, they also need to look at the scipy source to see how good 
an optimization algorithm can be it.
Sometimes they even have to hack the code to fit there needs (I did it...)

Scipy is so nice because it is powerfull and open source. All the users 
I know (except maybe one ;)) want to be able to compile the svn but they 
don't really like to play with system administration.

That is why I think we should have a very easy and accurate procedure to 
install scipy on ubuntu 8.10 (I could test on suse  if I have time).

Xavier

> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From xavier.gnata at gmail.com  Mon Oct  6 19:02:19 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Tue, 07 Oct 2008 01:02:19 +0200
Subject: [SciPy-user] Ubuntu Libraries
In-Reply-To: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>
References: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>
Message-ID: <48EA98FB.6010703@gmail.com>

jah wrote:
> Hi,
>
> I was hoping someone could clarify the ubuntu packages and what is 
> actually needed by scipy.
>
> 1) I keep seeing posts about g77 not being required anymore and 
> gfortran preferred.  What is the "official" statement?  In 
> INSTALL.txt, only g77 is mentioned for Ubuntu packages via:  
> """Debian/Ubuntu packages (g77): atlas3-base atlas3-base-dev""".  The 
> only reference to gfortran is with Mac OS X.  Also, in "Optional 
> Packages" debian packages 'gcc g++ g77' are recommened.
>
> 2) The install notes also say that a complete version of LAPACK is 
> required for scipy.  All the ubuntu descriptions say that only a 
> subset of routines from LAPACK are included with ATLAS.  For example, 
> http://packages.ubuntu.com/hardy/libatlas-sse2-dev.  Is this a problem?
>
> 3) It looks like there are two packages:  atlas3-sse2 and 
> libatlas-sse2-dev.  Are both required?  It seems that they refer to 
> different versions of ATLAS.
>
>  http://packages.ubuntu.com/hardy/atlas3-sse2
>  http://packages.ubuntu.com/hardy/libatlas-sse2-dev
>
> On reading http://www.scipy.org/Installing_SciPy/Linux, it seems like 
> I only should need libatlas-sse2-dev.  Correct?
>
> Thanks!

Well it really means that this documentation is not up to date or maybe 
not as accurate as it should be.
I'm trying to clarify that for the ubuntu 8.10 users (release end of 
october).

Xavier


From ondrej at certik.cz  Mon Oct  6 21:40:12 2008
From: ondrej at certik.cz (Ondrej Certik)
Date: Tue, 7 Oct 2008 03:40:12 +0200
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <48EA047A.2060504@ar.media.kyoto-u.ac.jp>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
	<48EA047A.2060504@ar.media.kyoto-u.ac.jp>
Message-ID: <85b5c3130810061840w333b4790k103adc8a07b1c6e9@mail.gmail.com>

On Mon, Oct 6, 2008 at 2:28 PM, David Cournapeau
<david at ar.media.kyoto-u.ac.jp> wrote:
> Xavier Gnata wrote:
>> Hi,
>>
>> I have installed an ubuntu 8.10 within a kvm image.
>> The goal is to write an as simple as possible procedure to compile
>> scipy on this distribution.
>> I would like to be able to compile scipy only after having installed
>> some packages. No tricks.
>> Good news, g77 is not needed anymore :)
>
> It was already the case for 8.04, but it was a bit confusing: both g77
> and gfortran libraries were available. Is it better with 8.10 ?
>
>>
>> I have one issue:
>> scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to
>> find 'umfpack.h'
>> I have installed libsuitesparse-dev providing /usr/include/suitesparse/umfpack.h
>> Of course, I can solve this with a symlink but it is very ugly.
>> It looks like there is something to fix in scipy auto-detection
>> because this "usr/include/suitesparse/umfpack.h" should be detected
>> and used.
>
> That's really the debian packagers' fault. Why do they think it is a
> good idea to change the header path of the library is beyond me; it
> breaks every single package which depends on it. That's stupid.
>
> We could get around it; but I though umfpack was being deprecated in
> scipy  (that is, we would do a scikit from it, but scipy would not
> depend on it anymore).

The way forward is to get involved with Debian packaging and get this fixed.

I did that for the scipy package and fixed that by applying a simple
patch to scipy. As to the default umfpack location, I also wondered
just like you, but it's useful to ask on the Debian list itself and
ask the people who do the packaging. :) So you can read the rationale
here (read the whole thread):

http://lists.alioth.debian.org/pipermail/pkg-scicomp-devel/2008-September/003133.html

I.e. citing:

"
we had a discussion with the author of suitesparse, the author of the
suitesparse interface in Octave and the person in charge of the fedora
octave/suitesparse package and we agreed in using /usr/include/suitesparse
as the place of the headers.
"

So if you (or anyone) have opinions on this, please join our teams in Debian:

http://wiki.debian.org/Teams/DebianScientificComputingTeam
http://wiki.debian.org/Teams/PythonModulesTeam

and let's get things fixed/discussed/moving.


As to building the official scipy/numpy packages in Debian or Ubuntu,
if something isn't working, it's my fault, as I did the uploads for
the last couple (a lot) revisions. Please report a bug in that case.



On Tue, Oct 7, 2008 at 12:59 AM, Xavier Gnata <xavier.gnata at gmail.com> wrote:
>>
>> That's really the debian packagers' fault. Why do they think it is a
>> good idea to change the header path of the library is beyond me; it
>> breaks every single package which depends on it. That's stupid.
>>
>>
> ok. I'm going to write a bug report. w&s.

Before you do, please read the thread I posted above.

Thanks,
Ondrej


From ondrej at certik.cz  Mon Oct  6 21:47:54 2008
From: ondrej at certik.cz (Ondrej Certik)
Date: Tue, 7 Oct 2008 03:47:54 +0200
Subject: [SciPy-user] Ubuntu Libraries
In-Reply-To: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>
References: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>
Message-ID: <85b5c3130810061847s5ffdc55agf3061e85ce09556c@mail.gmail.com>

On Mon, Oct 6, 2008 at 7:47 PM, jah <jah.mailinglist at gmail.com> wrote:
> Hi,
>
> I was hoping someone could clarify the ubuntu packages and what is actually
> needed by scipy.
>
> 1) I keep seeing posts about g77 not being required anymore and gfortran
> preferred.  What is the "official" statement?  In INSTALL.txt, only g77 is
> mentioned for Ubuntu packages via:  """Debian/Ubuntu packages (g77):
> atlas3-base atlas3-base-dev""".  The only reference to gfortran is with Mac
> OS X.  Also, in "Optional Packages" debian packages 'gcc g++ g77' are
> recommened.

Debian (and thus Ubuntu as well) has transitioned from g77 to
gfortran, so only gfortran is needed.

>
> 2) The install notes also say that a complete version of LAPACK is required
> for scipy.  All the ubuntu descriptions say that only a subset of routines
> from LAPACK are included with ATLAS.  For example,
> http://packages.ubuntu.com/hardy/libatlas-sse2-dev.  Is this a problem?

As far as I know, you can use atlas instead interchangebly. You can
find some (fixed) Debian bug reports about that in the python-numpy
package.

>
> 3) It looks like there are two packages:  atlas3-sse2 and
> libatlas-sse2-dev.  Are both required?  It seems that they refer to
> different versions of ATLAS.
>
>  http://packages.ubuntu.com/hardy/atlas3-sse2
>  http://packages.ubuntu.com/hardy/libatlas-sse2-dev
>
> On reading http://www.scipy.org/Installing_SciPy/Linux, it seems like I only
> should need libatlas-sse2-dev.  Correct?

The easiest way to determine the build dependencies is to do:

$ apt-get source python-scipy
$ cd python-scipy-0.6.0/
$ cat debian/control
Source: python-scipy
Section: python
Priority: extra
Maintainer: Debian Python Modules Team
<python-modules-team at lists.alioth.debian.org>
Uploaders: Alexandre Fayolle <afayolle at debian.org>, Marco Presi
(Zufus) <zufus at debian.org>, Ondrej Certik <ondrej at certik.cz>
Build-Depends: debhelper (>= 5.0.37.2), dpkg-dev (>= 1.13.19), quilt,
 python-all-dev, python-central (>= 0.5), python-numpy (>= 1:1.0.2), gfortran,
 sharutils, swig, libsuitesparse-dev (>= 3.1.0-3), libnetcdf-dev,
 libx11-dev, libblas-dev | libatlas-base-dev, liblapack-dev | libatlas-base-dev,
 libfftw3-dev
XS-Python-Version: all
Standards-Version: 3.8.0
Homepage: http://www.scipy.org/
Vcs-Svn: svn://svn.debian.org/python-modules/packages/scipy/trunk
Vcs-Browser: http://svn.debian.org/wsvn/python-modules/packages/scipy/trunk/?op=log
XS-DM-Upload-Allowed: yes


And you can read the Build-Depends right away.

Btw, is there any reason why the official Debian/Ubuntu packages are
not sufficient for you? I.e. is it because you need the svn version of
scipy? I haven't checked out if the svn scipy builds using the same
build dependencies as 0.6.0, that's true.

Ondrej


From peridot.faceted at gmail.com  Mon Oct  6 22:23:32 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Mon, 6 Oct 2008 22:23:32 -0400
Subject: [SciPy-user] Gaussian quadrature error
In-Reply-To: <BLU149-W65E2810654A9A934CB2428DC390@phx.gbl>
References: <0F8253EA348F49F6A3C3A0AEFE044866@GatewayLaptop>
	<ce557a360809170900r4f2105b6g24089a81a6612150@mail.gmail.com>
	<BLU149-W65E2810654A9A934CB2428DC390@phx.gbl>
Message-ID: <ce557a360810061923v60b64d2fpf7b4ca760239530f@mail.gmail.com>

2008/10/6 Dan Murphy <chiefmurph at hotmail.com>:
> Thanks for the clear explanation, Anne. If I understand the second take-home
> lesson -- feed integrators unity-order quantities -- then instead of calling
> quadrature as I did:
>    integrate.quadrature(f,-1000.0,0.0)
> I should scale my function f so that my call looks more like
>    integrate.quadrature(f,-1.0,0.0)
> Was that what you meant?

Yes, though with the particular function you used - exp(x) - you then
have the problem of exceedingly rapid changes within the region of
integration. If what you actually wanted was to integrate from
-infinity to zero, then you might do better to rescale your x
coordinate nonlinearly, say by x' = 1/(1-x) so that the integration
becomes one from zero to one without a sharp edge near one side.

A further take-home lesson is that Gaussian quadrature is extremely
efficient for functions that behave like high-order polynomials -
which exp(x) does on small intervals but not on large intervals.
However, you can use customized Gaussian quadrature based on a
different set of polynomials to integrate functions that look like a
polynomial times a weight factor.

> Also, when you say "scipy.optimize.quad is a little more general-purpose"
> did you mean "scipy.integrate.quad"? Indeed, the call
>    integrate(quad(f,-1000.0,0.0)
> worked great!

Oops. Yes. It's still possible to trick it, but scipy.integrate.quad
is based on old, well-tested code that is designed to be fairly robust
against peculiar functions. It will be a little slower than Gaussian
quadrature for functions that are extremely smooth, but then Gaussian
quadrature suffers very badly when handed functions like abs(x).

Anne


From peridot.faceted at gmail.com  Mon Oct  6 22:30:02 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Mon, 6 Oct 2008 22:30:02 -0400
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <bc657ead0810061159jd8d8b5cjf7f39a424c7a568d@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
	<5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>
	<9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>
	<ce557a360810061134s1fe2cfa5t288284e5d8315b69@mail.gmail.com>
	<bc657ead0810061159jd8d8b5cjf7f39a424c7a568d@mail.gmail.com>
Message-ID: <ce557a360810061930r2cff081hd700c16647ff3fc5@mail.gmail.com>

2008/10/6 Sebastian Haase <haase at msg.ucsf.edu>:
> On Mon, Oct 6, 2008 at 8:34 PM, Anne Archibald
> <peridot.faceted at gmail.com> wrote:
>> 2008/10/6 St?fan van der Walt <stefan at sun.ac.za>:
>>> 2008/9/26 David Cournapeau <cournape at gmail.com>:
>>>> Yes, it would be nice. What do other people think about deprecating
>>>> all the numpy re-export in scipy ? It would be nice to do for 0.7
>>>> (e.g. in 0.7, deprecated, in 0.8, removed).
>>>
>>> There were no objections to this, so may we go ahead?
>>
>> My only concern is possible user confusion: some functions (e.g. sqrt)
>> are provided as "enhanced" versions in scipy, while others are simply
>> reexported. If we remove the reexports, users can't simply use
>> scipy.whatever to get the best-available version of each function,
>> they have to know whether an enhanced version exists. Of course, since
>> the enhanced versions exist because their APIs differ in important and
>> possibly surprising ways (e.g., sqrt(-1) has a different return type
>> from sqrt(1)) this may be a good thing.
>>
> Arguing that SciPy is below 1.0 I think the reexporting should be
> minimized as much as possible.
> I don't thing that some people's preference for
> "from scipy import *"
> (without a preceding "from numpy import *") should not be a deciding point.

The case I was concerned about was

import scipy as sp
x = sp.cos(2*sp.arccos(y))

If this is changed to

import numpy as np
x = np.cos(2*np.arccos(y))

it suddenly stops working for values y>1. To keep the same behaviour
it needs to be

import scipy as sp
import numpy as np
x = np.cos(2*sp.arccos(y))

This is perhaps all right, but it does mean that users need to pay attention.

Anne

From robert.kern at gmail.com  Mon Oct  6 22:40:38 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 6 Oct 2008 21:40:38 -0500
Subject: [SciPy-user] Inconsistent standard deviation and variance
	implementation in scipy vs. scipy.stats
In-Reply-To: <ce557a360810061930r2cff081hd700c16647ff3fc5@mail.gmail.com>
References: <200809241605.01575.jr@sun.ac.za>
	<200809241205.02637.pgmdevlist@gmail.com>
	<48DA9519.1000203@american.edu>
	<ce557a360809241335w58e34486j8c6ed7d53549b2fc@mail.gmail.com>
	<5b8d13220809252119ya7b7b28p400d914792b129a8@mail.gmail.com>
	<9457e7c80810060928i239f2a7eg5dee5121617317b5@mail.gmail.com>
	<ce557a360810061134s1fe2cfa5t288284e5d8315b69@mail.gmail.com>
	<bc657ead0810061159jd8d8b5cjf7f39a424c7a568d@mail.gmail.com>
	<ce557a360810061930r2cff081hd700c16647ff3fc5@mail.gmail.com>
Message-ID: <3d375d730810061940q2d358482n80341e2367f8fd5c@mail.gmail.com>

On Mon, Oct 6, 2008 at 21:30, Anne Archibald <peridot.faceted at gmail.com> wrote:
> The case I was concerned about was
>
> import scipy as sp
> x = sp.cos(2*sp.arccos(y))
>
> If this is changed to
>
> import numpy as np
> x = np.cos(2*np.arccos(y))
>
> it suddenly stops working for values y>1. To keep the same behaviour
> it needs to be
>
> import scipy as sp
> import numpy as np
> x = np.cos(2*sp.arccos(y))
>
> This is perhaps all right, but it does mean that users need to pay attention.

Well, if we remove some names from scipy/__init__.py, we should remove
them all. The actual definitions of those extended-domain functions
are actually in numpy.lib.scimath, not scipy. We can make a convenient
module inside numpy that basically does this:

  from numpy import *
  from numpy.lib.scimath import *

Then the transition for people using "import scipy" becomes quite easy.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From david at ar.media.kyoto-u.ac.jp  Mon Oct  6 23:49:55 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 07 Oct 2008 12:49:55 +0900
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <85b5c3130810061840w333b4790k103adc8a07b1c6e9@mail.gmail.com>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>	<48EA047A.2060504@ar.media.kyoto-u.ac.jp>
	<85b5c3130810061840w333b4790k103adc8a07b1c6e9@mail.gmail.com>
Message-ID: <48EADC63.7020107@ar.media.kyoto-u.ac.jp>

Ondrej Certik wrote:
>
> The way forward is to get involved with Debian packaging and get this fixed.
>
> I did that for the scipy package and fixed that by applying a simple
> patch to scipy. As to the default umfpack location, I also wondered
> just like you, but it's useful to ask on the Debian list itself and
> ask the people who do the packaging. :) 

Oh, I know why they did that (avoiding cluttering /usr/include). I don't
think there is a point in discussing, they won't change it now. We have
to handle this in our umfpack detection scheme.

cheers,

David


From david at ar.media.kyoto-u.ac.jp  Mon Oct  6 23:53:38 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 07 Oct 2008 12:53:38 +0900
Subject: [SciPy-user] Ubuntu Libraries
In-Reply-To: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>
References: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>
Message-ID: <48EADD42.9080608@ar.media.kyoto-u.ac.jp>

jah wrote:
> Hi,
>
> I was hoping someone could clarify the ubuntu packages and what is
> actually needed by scipy.
>
> 1) I keep seeing posts about g77 not being required anymore and
> gfortran preferred.  What is the "official" statement?  In
> INSTALL.txt, only g77 is mentioned for Ubuntu packages via: 
> """Debian/Ubuntu packages (g77): atlas3-base atlas3-base-dev""".  The
> only reference to gfortran is with Mac OS X.  Also, in "Optional
> Packages" debian packages 'gcc g++ g77' are recommened.

It depends on the version. g77 and gfortran are not ABI compatible: you
can't mix them in your binary, including shared libraries. So the advice
was and still is to use the compiler of your distribution's ABI. Ubuntu
8.04 transitionned from g77 to gfortran ABI (both ABI were available in
e.g. atlas: atlas3-base-dev for g77 ABI vs libatlas-sse2-dev for
gfortran ABI).

>
> 2) The install notes also say that a complete version of LAPACK is
> required for scipy.  All the ubuntu descriptions say that only a
> subset of routines from LAPACK are included with ATLAS.  For example,
> http://packages.ubuntu.com/hardy/libatlas-sse2-dev.  Is this a problem?

Ubuntu (and debian) do include the full lapack into atlas: you have
nothing to do with them on Ubuntu.

>
> 3) It looks like there are two packages:  atlas3-sse2 and
> libatlas-sse2-dev.  Are both required?  It seems that they refer to
> different versions of ATLAS.
>
>  http://packages.ubuntu.com/hardy/atlas3-sse2
>  http://packages.ubuntu.com/hardy/libatlas-sse2-dev
>
> On reading http://www.scipy.org/Installing_SciPy/Linux, it seems like
> I only should need libatlas-sse2-dev.  Correct?

Yep, if you use gfortran. You should install libatlas-sse2-dev if you
build everything (numpy and scipy) with gfortran.

David


From jah.mailinglist at gmail.com  Tue Oct  7 03:23:11 2008
From: jah.mailinglist at gmail.com (jah)
Date: Tue, 7 Oct 2008 00:23:11 -0700
Subject: [SciPy-user] Ubuntu Libraries
In-Reply-To: <85b5c3130810061847s5ffdc55agf3061e85ce09556c@mail.gmail.com>
References: <df39e47d0810061047x2dac911fmfdc628c117384308@mail.gmail.com>
	<85b5c3130810061847s5ffdc55agf3061e85ce09556c@mail.gmail.com>
Message-ID: <df39e47d0810070023w212649adl79d4cdbc94753753@mail.gmail.com>

On Mon, Oct 6, 2008 at 6:47 PM, Ondrej Certik <ondrej at certik.cz> wrote:

>
> Btw, is there any reason why the official Debian/Ubuntu packages are
> not sufficient for you? I.e. is it because you need the svn version of
> scipy?


That is exactly why.  Also, started from scratch with Python 2.6, hoping to
learn something in the process.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081007/4950b9dc/attachment.html>

From xavier.gnata at gmail.com  Tue Oct  7 07:43:17 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Tue, 7 Oct 2008 13:43:17 +0200
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <48EADC63.7020107@ar.media.kyoto-u.ac.jp>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
	<48EA047A.2060504@ar.media.kyoto-u.ac.jp>
	<85b5c3130810061840w333b4790k103adc8a07b1c6e9@mail.gmail.com>
	<48EADC63.7020107@ar.media.kyoto-u.ac.jp>
Message-ID: <2a1f8a930810070443s70c45f69o4c98cdad5f63d473@mail.gmail.com>

On Tue, Oct 7, 2008 at 5:49 AM, David Cournapeau <
david at ar.media.kyoto-u.ac.jp> wrote:

> Ondrej Certik wrote:
> >
> > The way forward is to get involved with Debian packaging and get this
> fixed.
> >
> > I did that for the scipy package and fixed that by applying a simple
> > patch to scipy. As to the default umfpack location, I also wondered
> > just like you, but it's useful to ask on the Debian list itself and
> > ask the people who do the packaging. :)
>
> Oh, I know why they did that (avoiding cluttering /usr/include). I don't
> think there is a point in discussing, they won't change it now. We have
> to handle this in our umfpack detection scheme.
>
>
Thanks! Indeed it is fully pointless to try to change that. libsuitesparse
must have somthing special as a library but anyway...if it is fixed by
adding /usr/include/suitesparse in the scipy seearch path it is perfect :)

Ondrej : Maybe I should try to have a script tool able to take the sources
of the scipy package, replace scipy sources by the current svn ones and try
to rebuild the package every week or so.

Xavier
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081007/42331022/attachment.html>

From ondrej at certik.cz  Tue Oct  7 08:14:05 2008
From: ondrej at certik.cz (Ondrej Certik)
Date: Tue, 7 Oct 2008 14:14:05 +0200
Subject: [SciPy-user] Install scipy on ubnutu 8.10
In-Reply-To: <2a1f8a930810070443s70c45f69o4c98cdad5f63d473@mail.gmail.com>
References: <2a1f8a930810051201w24e83e04k1341884441c51aa@mail.gmail.com>
	<48EA047A.2060504@ar.media.kyoto-u.ac.jp>
	<85b5c3130810061840w333b4790k103adc8a07b1c6e9@mail.gmail.com>
	<48EADC63.7020107@ar.media.kyoto-u.ac.jp>
	<2a1f8a930810070443s70c45f69o4c98cdad5f63d473@mail.gmail.com>
Message-ID: <85b5c3130810070514q56cd8971l93bb170dc611c46d@mail.gmail.com>

On Tue, Oct 7, 2008 at 1:43 PM, Xavier Gnata <xavier.gnata at gmail.com> wrote:
>
>
> On Tue, Oct 7, 2008 at 5:49 AM, David Cournapeau
> <david at ar.media.kyoto-u.ac.jp> wrote:
>>
>> Ondrej Certik wrote:
>> >
>> > The way forward is to get involved with Debian packaging and get this
>> > fixed.
>> >
>> > I did that for the scipy package and fixed that by applying a simple
>> > patch to scipy. As to the default umfpack location, I also wondered
>> > just like you, but it's useful to ask on the Debian list itself and
>> > ask the people who do the packaging. :)
>>
>> Oh, I know why they did that (avoiding cluttering /usr/include). I don't
>> think there is a point in discussing, they won't change it now. We have
>> to handle this in our umfpack detection scheme.
>>
>
> Thanks! Indeed it is fully pointless to try to change that. libsuitesparse
> must have somthing special as a library but anyway...if it is fixed by
> adding /usr/include/suitesparse in the scipy seearch path it is perfect :)

Well, I thought it should change, but apparently there are reasons for
the current way. And for me, I don't really care where it is, as long
as it works. So imho adding the /usr/include/suitesparse in the search
path is the way to go.

>
> Ondrej : Maybe I should try to have a script tool able to take the sources
> of the scipy package, replace scipy sources by the current svn ones and try
> to rebuild the package every week or so.

Indeed, that'd be useful. Basically you just need the debian dir,
that's it. Let me know if you need anything fixed. As I said, if you
have time, we'd be happy if you join the Debian Python Modules Team
and help us maintain the scipy/numpy packages in Debian/Ubuntu.

Ondrej


From nmarais at sun.ac.za  Tue Oct  7 15:06:54 2008
From: nmarais at sun.ac.za (Neilen Marais)
Date: Tue, 7 Oct 2008 19:06:54 +0000 (UTC)
Subject: [SciPy-user] Solving complex RHS for real matrix using umfpack and
 scipy.sparse.linalg.dsolve.factorized
Message-ID: <gcgc0e$dh5$1@ger.gmane.org>

Hi,

I have a real sparse matrix that I factorized using 
scipy.sparse.linalg.dsolve.factorized(). When I solve it with a complex 
RHS, I always get a real return. Do I need to set the matrix type as 
complex in this case, or is there a better way?

Thanks
Neilen



From peridot.faceted at gmail.com  Tue Oct  7 16:12:37 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Tue, 7 Oct 2008 16:12:37 -0400
Subject: [SciPy-user] Solving complex RHS for real matrix using umfpack
	and scipy.sparse.linalg.dsolve.factorized
In-Reply-To: <gcgc0e$dh5$1@ger.gmane.org>
References: <gcgc0e$dh5$1@ger.gmane.org>
Message-ID: <ce557a360810071312i16e56716k1e5ae00decbdc56f@mail.gmail.com>

2008/10/7 Neilen Marais <nmarais at sun.ac.za>:

> I have a real sparse matrix that I factorized using
> scipy.sparse.linalg.dsolve.factorized(). When I solve it with a complex
> RHS, I always get a real return. Do I need to set the matrix type as
> complex in this case, or is there a better way?

If all you're doing is solving  y = A*x, then you can simply solve for
the real and imaginary parts separately, since a real matrix won't mix
them and the problem is linear.

Anne


From nwagner at iam.uni-stuttgart.de  Wed Oct  8 10:39:51 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 08 Oct 2008 16:39:51 +0200
Subject: [SciPy-user] timeseries documentation via sphinx
Message-ID: <web-112500515@uni-stuttgart.de>

Hi all,

I tried to build the documentation of timeseries

make latex PAPER=a4
mkdir -p build/latex build/doctrees
/data/home/nwagner/local/bin/sphinx-build -b latex -d 
build/doctrees -D latex_paper_size=a4  source build/latex
/data/home/nwagner/local/lib/python2.5/site-packages/matplotlib/__init__.py:367: 
UserWarning: matplotlibrc text.usetex can not be used with 
*Agg backend unless dvipng-1.5 or later is installed on 
your system
   warnings.warn( 'matplotlibrc text.usetex can not be 
used with *Agg '
Sphinx v0.4.3, building latex
trying to load pickled env... not found
Exception occurred:
   File 
"/data/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/../ext/numpydoc.py", 
line 361, in monkeypatch_sphinx_ext_autodoc
     if autodoc.RstGenerator.format_signature is 
our_format_signature:
AttributeError: 'module' object has no attribute 
'RstGenerator'
The full traceback has been saved in 
/tmp/sphinx-err-2jMNGx.log, if you want to report the 
issue to the author.
Please also report this if it was a user error, so that a 
better error message can be provided next time.
Send reports to sphinx-dev at googlegroups.com. Thanks!
make: *** [latex] Fehler 1
  
How can I resolve this problem ?

Nils


From david at ar.media.kyoto-u.ac.jp  Wed Oct  8 10:28:17 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 08 Oct 2008 23:28:17 +0900
Subject: [SciPy-user] timeseries documentation via sphinx
In-Reply-To: <web-112500515@uni-stuttgart.de>
References: <web-112500515@uni-stuttgart.de>
Message-ID: <48ECC381.6040403@ar.media.kyoto-u.ac.jp>

Nils Wagner wrote:
> Hi all,
>
> I tried to build the documentation of timeseries
>   

You need the developement version of sphinx (0.5dev)

David


From philbinj at gmail.com  Wed Oct  8 11:41:48 2008
From: philbinj at gmail.com (James Philbin)
Date: Wed, 8 Oct 2008 16:41:48 +0100
Subject: [SciPy-user] Left hand sparse matrix multiplication
Message-ID: <2b1c8c4f0810080841l4df18c1fre07ee259a6164f74@mail.gmail.com>

Hi,

I'm trying to compute x*A where x is a dense row vector and A is a
sparse CSC matrix. A.rmatvec seems to do what I want but is wasteful
as it computes:
self.transpose().matvec( other )
i.e. it computes A^T * x^T.

It seems there should be a much more efficient overload for csc's
rmatvec which doesn't involve computing the transpose. I hope i'm
understanding things correctly.

Thanks,
James


On Tue, Oct 7, 2008 at 9:12 PM, Anne Archibald
<peridot.faceted at gmail.com> wrote:
> 2008/10/7 Neilen Marais <nmarais at sun.ac.za>:
>
>> I have a real sparse matrix that I factorized using
>> scipy.sparse.linalg.dsolve.factorized(). When I solve it with a complex
>> RHS, I always get a real return. Do I need to set the matrix type as
>> complex in this case, or is there a better way?
>
> If all you're doing is solving  y = A*x, then you can simply solve for
> the real and imaginary parts separately, since a real matrix won't mix
> them and the problem is linear.
>
> Anne
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From nwagner at iam.uni-stuttgart.de  Wed Oct  8 12:32:14 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 08 Oct 2008 18:32:14 +0200
Subject: [SciPy-user] timeseries documentation via sphinx
In-Reply-To: <48ECC381.6040403@ar.media.kyoto-u.ac.jp>
References: <web-112500515@uni-stuttgart.de>
	<48ECC381.6040403@ar.media.kyoto-u.ac.jp>
Message-ID: <web-112506740@uni-stuttgart.de>

On Wed, 08 Oct 2008 23:28:17 +0900
  David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
> Nils Wagner wrote:
>> Hi all,
>>
>> I tried to build the documentation of timeseries
>>   
> 
> You need the developement version of sphinx (0.5dev)
> 
> David

Hi David,

Do I need more than

svn co 
http://svn.python.org/projects/doctools/trunk/sphinx 
sphinx
  
Nils


From mattknox.ca at gmail.com  Wed Oct  8 12:33:14 2008
From: mattknox.ca at gmail.com (Matt Knox)
Date: Wed, 8 Oct 2008 16:33:14 +0000 (UTC)
Subject: [SciPy-user] timeseries documentation via sphinx
References: <web-112500515@uni-stuttgart.de>
Message-ID: <loom.20081008T162808-331@post.gmane.org>

> Hi all,
> 
> I tried to build the documentation of timeseries
> ......
> 
> How can I resolve this problem ?
> 
> Nils
> 

Also note that there are pre-built docs at http://pytseries.sourceforge.net/
If you have trouble building them after switching to the development version of
sphinx, let me know.

- Matt



From dominique.orban at gmail.com  Wed Oct  8 12:36:10 2008
From: dominique.orban at gmail.com (Dominique Orban)
Date: Wed, 8 Oct 2008 12:36:10 -0400
Subject: [SciPy-user] Left hand sparse matrix multiplication
In-Reply-To: <2b1c8c4f0810080841l4df18c1fre07ee259a6164f74@mail.gmail.com>
References: <2b1c8c4f0810080841l4df18c1fre07ee259a6164f74@mail.gmail.com>
Message-ID: <8793ae6e0810080936m3a2b65f5vcce043968f5f8f8e@mail.gmail.com>

On Wed, Oct 8, 2008 at 11:41 AM, James Philbin <philbinj at gmail.com> wrote:
> Hi,
>
> I'm trying to compute x*A where x is a dense row vector and A is a
> sparse CSC matrix. A.rmatvec seems to do what I want but is wasteful
> as it computes:
> self.transpose().matvec( other )
> i.e. it computes A^T * x^T.
>
> It seems there should be a much more efficient overload for csc's
> rmatvec which doesn't involve computing the transpose. I hope i'm
> understanding things correctly.

Do you only need matrix-vector products with A in this form, i.e.,
x*A, or do you also need A*x? If you only need x*A you're probably
better off storing B=A^T in CSR format and computing B*x' where x' is
the column vector x (always stored as a column vector.)

Dominique


> On Tue, Oct 7, 2008 at 9:12 PM, Anne Archibald
> <peridot.faceted at gmail.com> wrote:
>> 2008/10/7 Neilen Marais <nmarais at sun.ac.za>:
>>
>>> I have a real sparse matrix that I factorized using
>>> scipy.sparse.linalg.dsolve.factorized(). When I solve it with a complex
>>> RHS, I always get a real return. Do I need to set the matrix type as
>>> complex in this case, or is there a better way?
>>
>> If all you're doing is solving  y = A*x, then you can simply solve for
>> the real and imaginary parts separately, since a real matrix won't mix
>> them and the problem is linear.
>>
>> Anne
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From nwagner at iam.uni-stuttgart.de  Wed Oct  8 13:20:03 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 08 Oct 2008 19:20:03 +0200
Subject: [SciPy-user] timeseries documentation via sphinx
In-Reply-To: <48ECC381.6040403@ar.media.kyoto-u.ac.jp>
References: <web-112500515@uni-stuttgart.de>
	<48ECC381.6040403@ar.media.kyoto-u.ac.jp>
Message-ID: <web-112508197@uni-stuttgart.de>

On Wed, 08 Oct 2008 23:28:17 +0900
  David Cournapeau <david at ar.media.kyoto-u.ac.jp> wrote:
> Nils Wagner wrote:
>> Hi all,
>>
>> I tried to build the documentation of timeseries
>>   
> 
> You need the developement version of sphinx (0.5dev)
> 
> David

I have updated to the development version of sphinx ...

make latex
mkdir -p build/latex build/doctrees
sphinx-build -b latex -d build/doctrees -D 
latex_paper_size=a4  source build/latex
Sphinx v0.5, building latex
loading pickled environment... not found
building [latex]: all documents
updating environment: 14 added, 0 changed, 0 removed
reading sources... core/Date core/DateArray 
core/TimeSeries core/index index installing intro 
lib/database lib/filtering lib/index lib/interpolation 
lib/plotting DEBUG: current directory: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc
DEBUG: 
fullpath:/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/yahoo.py
DEBUG: 
outdirnm:/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/yahoo.py

/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/yahoo.png
     building 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/yahoo.py
DEBUG: current directory: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc
DEBUG: 
fullpath:/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/expmave.py
DEBUG: 
outdirnm:/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/expmave.py

/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/expmave.png
     building 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/expmave.py
DEBUG: current directory: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc
DEBUG: 
fullpath:/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/sepaxis.py
DEBUG: 
outdirnm:/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/sepaxis.py

/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/sepaxis.png
     building 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/sepaxis.py
DEBUG: current directory: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc
DEBUG: 
fullpath:/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom1.py
DEBUG: 
outdirnm:/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom1.py

/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom1.png
     building 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom1.py
DEBUG: current directory: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc
DEBUG: 
fullpath:/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom2.py
DEBUG: 
outdirnm:/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom2.py

/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom2.png
     building 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom2.py
DEBUG: current directory: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc
DEBUG: 
fullpath:/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom3.py
DEBUG: 
outdirnm:/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom3.py

/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom3.png
     building 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom3.py
DEBUG: current directory: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc
DEBUG: 
fullpath:/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom4.py
DEBUG: 
outdirnm:/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom4.py

/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom4.png
     building 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/plotting/zoom4.py
lib/report license
WARNING: 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/source/lib/database.rst:132: 
(WARNING/2) autodoc can't import/find module 
'scikits.timeseries.lib.tstables', it reported error: "No 
module named tables", please check your spelling and 
sys.path
pickling environment... done
checking consistency... done
processing TimeSeries.tex... index intro license 
installing core/index core/Date core/DateArray 
core/TimeSeries lib/index lib/interpolation lib/filtering 
lib/plotting lib/report lib/database
resolving references...
writing... Exception occurred:
   File 
"/usr/local/lib64/python2.5/site-packages/Sphinx-0.5dev_20081008-py2.5.egg/sphinx/latexwriter.py", 
line 580, in visit_entry
     raise NotImplementedError('Column or row spanning 
cells are '
NotImplementedError: Column or row spanning cells are not 
implemented.
The full traceback has been saved in 
/tmp/sphinx-err-33YnJ3.log, if you want to report the 
issue to the author.
Please also report this if it was a user error, so that a 
better error message can be provided next time.
Send reports to sphinx-dev at googlegroups.com. Thanks!
make: *** [latex] Error 1

Nils

  


From pgmdevlist at gmail.com  Wed Oct  8 13:20:35 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Wed, 8 Oct 2008 13:20:35 -0400
Subject: [SciPy-user] timeseries documentation via sphinx
In-Reply-To: <web-112508197@uni-stuttgart.de>
References: <web-112500515@uni-stuttgart.de>
	<48ECC381.6040403@ar.media.kyoto-u.ac.jp>
	<web-112508197@uni-stuttgart.de>
Message-ID: <200810081320.35661.pgmdevlist@gmail.com>

Oh, looks like somebody committed some fixes without removing the DEBUG 
statements. Sorry about that.
Anyhow: yes, there's a problem with the tables. I'll try to find a workaround. 
In the meantime, the html docs should work OK.


From nwagner at iam.uni-stuttgart.de  Wed Oct  8 13:33:35 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 08 Oct 2008 19:33:35 +0200
Subject: [SciPy-user] timeseries documentation via sphinx
In-Reply-To: <200810081320.35661.pgmdevlist@gmail.com>
References: <web-112500515@uni-stuttgart.de>
	<48ECC381.6040403@ar.media.kyoto-u.ac.jp>
	<web-112508197@uni-stuttgart.de>
	<200810081320.35661.pgmdevlist@gmail.com>
Message-ID: <web-112508545@uni-stuttgart.de>

On Wed, 8 Oct 2008 13:20:35 -0400
  Pierre GM <pgmdevlist at gmail.com> wrote:
> Oh, looks like somebody committed some fixes without 
>removing the DEBUG 
> statements. Sorry about that.
> Anyhow: yes, there's a problem with the tables. I'll try 
>to find a workaround. 
> In the meantime, the html docs should work OK.

Yes, works for me ! Thank you very much !
make html
mkdir -p build/html build/doctrees
sphinx-build -b html -d build/doctrees -D 
latex_paper_size=a4  source build/html
Sphinx v0.5, building html
loading pickled environment... done
building [html]: targets for 14 source files that are out 
of date
updating environment: 0 added, 1 changed, 0 removed
reading sources... core/TimeSeries
pickling environment... done
checking consistency... done
preparing documents... done
writing output... core/Date core/DateArray core/TimeSeries 
core/index index installing intro lib/database 
lib/filtering lib/index lib/interpolation lib/plotting 
lib/report license
writing additional files... genindex search
copying images... 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom3.png 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom2.png 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/expmave.png 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom1.pdf 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/yahoo.pdf 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom4.png 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom1.png 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/sepaxis.png 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/sepaxis.pdf 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom4.pdf 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom2.pdf 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/yahoo.png 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/expmave.pdf 
/home/nwagner/svn/timeseries/scikits/timeseries/doc/build/plots/zoom3.pdf
copying static files... done
dumping search index... done
dumping object inventory... done
build succeeded.

Build finished. The HTML pages are in build/html.
  


From wnbell at gmail.com  Wed Oct  8 15:28:05 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Wed, 8 Oct 2008 15:28:05 -0400
Subject: [SciPy-user] Left hand sparse matrix multiplication
In-Reply-To: <2b1c8c4f0810080841l4df18c1fre07ee259a6164f74@mail.gmail.com>
References: <2b1c8c4f0810080841l4df18c1fre07ee259a6164f74@mail.gmail.com>
Message-ID: <d05265cb0810081228r47731877ic187039ec71c0b6b@mail.gmail.com>

On Wed, Oct 8, 2008 at 11:41 AM, James Philbin <philbinj at gmail.com> wrote:
>
> I'm trying to compute x*A where x is a dense row vector and A is a
> sparse CSC matrix. A.rmatvec seems to do what I want but is wasteful
> as it computes:
> self.transpose().matvec( other )
> i.e. it computes A^T * x^T.
>
> It seems there should be a much more efficient overload for csc's
> rmatvec which doesn't involve computing the transpose. I hope i'm
> understanding things correctly.
>

CSR.T and CSC.T are constant time operations, they just return the
matrix in the "opposite" format.

In your case, A.T is equivalent to
csr_matrix((A.data,A.indices,A.indptr),
shape=(A.shape[1],A.shape[0])), which simply reinterprets the CSC
format of A as the CSR format of A.T.

This does not hold for other sparse formats so there *is* some room
for improvement.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From Dharhas.Pothina at twdb.state.tx.us  Wed Oct  8 15:54:30 2008
From: Dharhas.Pothina at twdb.state.tx.us (Dharhas Pothina)
Date: Wed, 08 Oct 2008 14:54:30 -0500
Subject: [SciPy-user] Calculating daily,
	monthly and seasonal averages of hourly	time series data.
Message-ID: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>

Hi,

I'm trying to analyze hourly salinity data. I was wondering if there is a simple way of calculating daily, monthly and seasonal averages of hourly time series data. 

So assuming I have two arrays that contain several years of hourly (or every 15min) salinity data: a datetime array called 'fielddates' & a data array called 'salinity'

How would I go about getting the various averages. The seasonal averages would be say defined as May through September etc. 

I had a look at scikits.timeseries but it looks like it would require upgrading numpy to install and there isn't enough high level documentation on how to use it for me to be confident in picking it up in the time frame I'm looking at. I'm also not completely clear if it can handle stuff that happens on a scale smaller than a day. If anyone can point me to any usage examples for it that would be appreciated.

Thanks,

- dharhas



From lroubeyrie at limair.asso.fr  Thu Oct  9 04:51:27 2008
From: lroubeyrie at limair.asso.fr (Lionel Roubeyrie)
Date: Thu, 09 Oct 2008 10:51:27 +0200
Subject: [SciPy-user] Calculating daily,
	monthly and seasonal averages	of hourly	time series data.
In-Reply-To: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>
References: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>
Message-ID: <1223542287.26062.20.camel@poste5>

Hi Dharhas,
scikits.timeseries is perfect for what you want in a very useable way :

###############################
In [29]: import scikits.timeseries as ts

In [30]: sdate=ts.Date('H', '2007-01-01 00:00')

In [31]: fielddates=ts.date_array(start_date=sdate, freq='H',
length=365*24*2)

In [32]: salinity=random(365*24*2)*100

In [33]: mes=ts.time_series(data=salinity, dates=fielddates)

In [34]: mes
Out[34]: 
timeseries([ 23.84116045  49.51437251  89.29221711 ...,  37.00510947
41.12589836
  78.65572656],
           dates = [01-jan-2007 00:00 ... 30-d?c-2008 23:00],
           freq  = H)


In [35]: mes_avmonth=mes.convert(freq='M', func=mean)

In [36]: mes_avmonth
Out[36]: 
timeseries([ 49.29718906  50.64688937  49.88193999  48.97144253
49.5788259
  50.41340038  50.15047009  51.70933261  50.5635153   51.15084406
  51.15362514  51.51443468  49.17556599  49.26877667  50.21416724
  49.37037657  51.00724033  49.43337134  49.60398056  50.24470761
  50.62350109  51.15572702  51.37652011  49.24193747],
           dates = [jan-2007 ... d?c-2008],
           freq  = M)


In [37]: mes_avyear=mes.convert(freq='Y', func=mean)

In [38]: mes_avyear
Out[38]: 
timeseries([ 50.41903159  50.06468157],
           dates = [2007 2008],
           freq  = A-DEC)


In [39]: mes_avseason=mes[(mes.month>=5) & (mes.month<=9)].mean()

In [40]: mes_avseason
Out[40]: 50,33380690600049
###############################


Le mercredi 08 octobre 2008 ? 14:54 -0500, Dharhas Pothina a ?crit :
> Hi,
> 
> I'm trying to analyze hourly salinity data. I was wondering if there is a simple way of calculating daily, monthly and seasonal averages of hourly time series data. 
> 
> So assuming I have two arrays that contain several years of hourly (or every 15min) salinity data: a datetime array called 'fielddates' & a data array called 'salinity'
> 
> How would I go about getting the various averages. The seasonal averages would be say defined as May through September etc. 
> 
> I had a look at scikits.timeseries but it looks like it would require upgrading numpy to install and there isn't enough high level documentation on how to use it for me to be confident in picking it up in the time frame I'm looking at. I'm also not completely clear if it can handle stuff that happens on a scale smaller than a day. If anyone can point me to any usage examples for it that would be appreciated.
> 
> Thanks,
> 
> - dharhas
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
> 
-- 
Lionel Roubeyrie
charg? d'?tudes
LIMAIR - La Surveillance de l'Air en Limousin
http://www.limair.asso.fr




From icy.flame.gm at gmail.com  Thu Oct  9 07:44:16 2008
From: icy.flame.gm at gmail.com (iCy-fLaME)
Date: Thu, 9 Oct 2008 12:44:16 +0100
Subject: [SciPy-user] Extrema finding
Message-ID: <bfb379e10810090444i4e435eb8q5f5729005087881b@mail.gmail.com>

I am trying to find a list of all maxima and minima, each for a given
(1D) numpy array. Anyone know of a quick way to do it?

Ideally the function will return the extrema values and their
positions. Relatively simple function to implement in Python, but that
would be painfully slow. The typical data array I am looking at, has
approximately 500k elements of double precision float.

Any thoughts and suggestions are much appreciated. Thanks!


From ndbecker2 at gmail.com  Thu Oct  9 08:06:06 2008
From: ndbecker2 at gmail.com (Neal Becker)
Date: Thu, 09 Oct 2008 08:06:06 -0400
Subject: [SciPy-user] Extrema finding
References: <bfb379e10810090444i4e435eb8q5f5729005087881b@mail.gmail.com>
Message-ID: <gcks3e$7bt$2@ger.gmane.org>

iCy-fLaME wrote:

> I am trying to find a list of all maxima and minima, each for a given
> (1D) numpy array. Anyone know of a quick way to do it?
> 
> Ideally the function will return the extrema values and their
> positions. Relatively simple function to implement in Python, but that
> would be painfully slow. The typical data array I am looking at, has
> approximately 500k elements of double precision float.
> 
> Any thoughts and suggestions are much appreciated. Thanks!

I have code for that in c++, which can be used with pyublas and boost::python. :)



From jdh2358 at gmail.com  Thu Oct  9 08:21:47 2008
From: jdh2358 at gmail.com (John Hunter)
Date: Thu, 9 Oct 2008 07:21:47 -0500
Subject: [SciPy-user] Extrema finding
In-Reply-To: <bfb379e10810090444i4e435eb8q5f5729005087881b@mail.gmail.com>
References: <bfb379e10810090444i4e435eb8q5f5729005087881b@mail.gmail.com>
Message-ID: <88e473830810090521qe247f41l9649f3c460e95d38@mail.gmail.com>

On Thu, Oct 9, 2008 at 6:44 AM, iCy-fLaME <icy.flame.gm at gmail.com> wrote:
> I am trying to find a list of all maxima and minima, each for a given
> (1D) numpy array. Anyone know of a quick way to do it?


>
> Ideally the function will return the extrema values and their
> positions. Relatively simple function to implement in Python, but that
> would be painfully slow. The typical data array I am looking at, has
> approximately 500k elements of double precision float.

I use the following to find the indices of extrema:


def extrema_sampling(x, withend=False):
    """
    return the indices into the local maxima or minima of x
    if withend, include the 0 and end points in the sampling
    """
    d = np.diff(x)
    d2 = d[:-1]*d[1:]
    ind = []
    if withend:
        ind.append(0)
    ind.extend(np.nonzero(d2<0)[0]+1)
    if withend and ind[-1]!=len(x)-1:
        ind.append(len(x)-1)
    return np.array(ind)


From dan.collins at uchsc.edu  Thu Oct  9 10:33:19 2008
From: dan.collins at uchsc.edu (dan collins uchsc)
Date: Thu, 9 Oct 2008 08:33:19 -0600
Subject: [SciPy-user] Extrema finding
Message-ID: <df738a0a0810090733ic2a62d1hbba2ac8432638f0f@mail.gmail.com>

I use the following technique to fine minima and maxima.
x = 1d vector
y = x; y = delete(y, [0], axis=0); y = append(y,0)
xy = x-y
a = [0]
for i in range(0, size(xy)):
    a = append(a,xy[i]-xy[i+1])

minima = where(a == 2)
maximu = where(a == -2)

>
>
>
> Message: 7
> Date: Thu, 9 Oct 2008 12:44:16 +0100
> From: iCy-fLaME <icy.flame.gm at gmail.com>
> Subject: [SciPy-user] Extrema finding
> To: "SciPy Users List" <scipy-user at scipy.org>
> Message-ID:
>        <bfb379e10810090444i4e435eb8q5f5729005087881b at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> I am trying to find a list of all maxima and minima, each for a given
> (1D) numpy array. Anyone know of a quick way to do it?
>
> Ideally the function will return the extrema values and their
> positions. Relatively simple function to implement in Python, but that
> would be painfully slow. The typical data array I am looking at, has
> approximately 500k elements of double precision float.
>
> Any thoughts and suggestions are much appreciated. Thanks!
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081009/e8beed44/attachment.html>

From lroubeyrie at limair.asso.fr  Thu Oct  9 10:39:49 2008
From: lroubeyrie at limair.asso.fr (Lionel Roubeyrie)
Date: Thu, 09 Oct 2008 16:39:49 +0200
Subject: [SciPy-user] Calculating daily,
	monthly and seasonal	averages	of hourly	time series data.
In-Reply-To: <48EDB6AF.63BA.009B.0@twdb.state.tx.us>
References: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>
	<1223542287.26062.20.camel@poste5>
	<48EDB6AF.63BA.009B.0@twdb.state.tx.us>
Message-ID: <1223563189.30896.47.camel@poste5>


> Ok I think I understood your example below. Can you give me an example
> of how to deal with missing data?
If you take my last example, you have 3 separated arrays:
mes.data : your values
mes.dates : the date array
mes.mask : like the maskedarray module (timeseries is based on it).
Trying to mask the 10 first values:
##################################
In [16]: mask=zeros_like(mes.data)

In [17]: mask[0:10]=True

In [18]: mask
Out[18]: array([ 1.,  1.,  1., ...,  0.,  0.,  0.])

In [19]: mes2=ts.time_series(mes, mask=mask)

In [20]: mes2
Out[20]: 
timeseries([-- -- -- ..., 17.9699245692 66.8968405206 24.7117965045],
           dates = [01-jan-2007 00:00 ... 30-d?c-2008 23:00],
           freq  = H)
##################################
A timeseries can be constructed based on another timeseries, like I do
here with mes2. Note that just the values are masked (missing), not the
dates because all fields have a value (masked or not).


>  Does this general technique work for data that is on a 15 minute frequency 
Yes, but no :-) The timeseries module doesn't handle directly QH
frequency, but minute frequency (freq='T'). Look at that :
#################################
In [28]: fielddates=ts.date_array(['2007-01-01 00:00', '2007-01-01
00:15', '2007-01-01 00:30', '2007-01-01 00:45'], freq='T')

In [29]: salinity=random(4)*100

In [30]: mes=ts.time_series(data=salinity, dates=fielddates)

In [31]: mes.has_missing_dates()
Out[31]: True
#################################
There's not QH native frequency, then there's some missing dates (you
can also look for duplicated dates, very convenient!). But you can fill
these missing dates :
###############################
In [36]: mes2=mes.fill_missing_dates()

In [37]: mes2.has_missing_dates()
Out[37]: False

In [38]: mes2
Out[38]: 
timeseries([2.33824586442 -- -- -- -- -- -- -- -- -- -- -- -- -- --
36.180901427 -- --
 -- -- -- -- -- -- -- -- -- -- -- -- 39.0648471531 -- -- -- -- -- -- --
--
 -- -- -- -- -- -- 55.4226606997],
           dates = [01-jan-2007 00:00 ... 01-jan-2007 00:45],
           freq  = T)
###############################
Or the module can handle directly these missing dates when you convert
the timeseries to a lower frequency:
###############################
In [39]: mes.convert(freq='H', func=mean)
Out[39]: 
timeseries([ 33.25166379],
           dates = [01-jan-2007 00:00],
           freq  = H)
###############################
You can try with func=None, it will just fill the missing dates with
missing values :-p

> or datasets where the frequency is
> variable (ie some months we have 10 readings other months we may have
> 30?
Like you see, just pass you datas with the corrects dates, and it rocks,
but don't mix minute frequency with hour frequency!
Here I take 3 daily samples in january, and one in october :
#################################
In [41]: fielddates=ts.date_array(['2007-01-01', '2007-01-02',
'2007-01-03', '2007-10-15'], freq='D')

In [42]: salinity=random(4)*100

In [43]: mes=ts.time_series(data=salinity, dates=fielddates)

In [44]: mes
Out[44]: 
timeseries([ 59.63468614  38.60721076  64.52554805  66.17637291],
           dates = [01-jan-2007 02-jan-2007 03-jan-2007 15-oct-2007],
           freq  = D)


In [45]: mes.convert(freq='M', func=mean)
Out[45]: 
timeseries([54.2558149823 -- -- -- -- -- -- -- -- 66.1763729106],
           dates = [jan-2007 ... oct-2007],
           freq  = M)
###################################
Computing the monthly average goes fine, the module fill the missing
months by masked values.

> 
> Also how stable is the scikits.timeseries? Is it reasonably usable?
Yes, we use it intensively on large projects and Pierre G.M. has made a
very good tool.
Cordialy

> 
> thanks,
> 
> - dharhas
> 
> >>> Lionel Roubeyrie <lroubeyrie at limair.asso.fr> 10/9/2008 3:51 AM >>>
> Hi Dharhas,
> scikits.timeseries is perfect for what you want in a very useable way
> :
> 
> ###############################
> In [29]: import scikits.timeseries as ts
> 
> In [30]: sdate=ts.Date('H', '2007-01-01 00:00')
> 
> In [31]: fielddates=ts.date_array(start_date=sdate, freq='H',
> length=365*24*2)
> 
> In [32]: salinity=random(365*24*2)*100
> 
> In [33]: mes=ts.time_series(data=salinity, dates=fielddates)
> 
> In [34]: mes
> Out[34]: 
> timeseries([ 23.84116045  49.51437251  89.29221711 ...,  37.00510947
> 41.12589836
>   78.65572656],
>            dates = [01-jan-2007 00:00 ... 30-d?c-2008 23:00],
>            freq  = H)
> 
> 
> In [35]: mes_avmonth=mes.convert(freq='M', func=mean)
> 
> In [36]: mes_avmonth
> Out[36]: 
> timeseries([ 49.29718906  50.64688937  49.88193999  48.97144253
> 49.5788259
>   50.41340038  50.15047009  51.70933261  50.5635153   51.15084406
>   51.15362514  51.51443468  49.17556599  49.26877667  50.21416724
>   49.37037657  51.00724033  49.43337134  49.60398056  50.24470761
>   50.62350109  51.15572702  51.37652011  49.24193747],
>            dates = [jan-2007 ... d?c-2008],
>            freq  = M)
> 
> 
> In [37]: mes_avyear=mes.convert(freq='Y', func=mean)
> 
> In [38]: mes_avyear
> Out[38]: 
> timeseries([ 50.41903159  50.06468157],
>            dates = [2007 2008],
>            freq  = A-DEC)
> 
> 
> In [39]: mes_avseason=mes[(mes.month>=5) & (mes.month<=9)].mean()
> 
> In [40]: mes_avseason
> Out[40]: 50,33380690600049
> ###############################
> 
> 
> Le mercredi 08 octobre 2008 ? 14:54 -0500, Dharhas Pothina a ?crit :
> > Hi,
> > 
> > I'm trying to analyze hourly salinity data. I was wondering if there
> is a simple way of calculating daily, monthly and seasonal averages of
> hourly time series data. 
> > 
> > So assuming I have two arrays that contain several years of hourly
> (or every 15min) salinity data: a datetime array called 'fielddates' & a
> data array called 'salinity'
> > 
> > How would I go about getting the various averages. The seasonal
> averages would be say defined as May through September etc. 
> > 
> > I had a look at scikits.timeseries but it looks like it would require
> upgrading numpy to install and there isn't enough high level
> documentation on how to use it for me to be confident in picking it up
> in the time frame I'm looking at. I'm also not completely clear if it
> can handle stuff that happens on a scale smaller than a day. If anyone
> can point me to any usage examples for it that would be appreciated.
> > 
> > Thanks,
> > 
> > - dharhas
> > 
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org 
> > http://projects.scipy.org/mailman/listinfo/scipy-user 
> > 
-- 
Lionel Roubeyrie
charg? d'?tudes
LIMAIR - La Surveillance de l'Air en Limousin
http://www.limair.asso.fr




From icy.flame.gm at gmail.com  Thu Oct  9 12:20:08 2008
From: icy.flame.gm at gmail.com (iCy-fLaME)
Date: Thu, 9 Oct 2008 17:20:08 +0100
Subject: [SciPy-user] Extrema finding
In-Reply-To: <88e473830810090521qe247f41l9649f3c460e95d38@mail.gmail.com>
References: <bfb379e10810090444i4e435eb8q5f5729005087881b@mail.gmail.com>
	<88e473830810090521qe247f41l9649f3c460e95d38@mail.gmail.com>
Message-ID: <bfb379e10810090920i290af59dtfdf2c1394ebc4b1f@mail.gmail.com>

On Thu, Oct 9, 2008 at 1:21 PM, John Hunter <jdh2358 at gmail.com> wrote:
> On Thu, Oct 9, 2008 at 6:44 AM, iCy-fLaME <icy.flame.gm at gmail.com> wrote:
>> I am trying to find a list of all maxima and minima, each for a given
>> (1D) numpy array. Anyone know of a quick way to do it?
>
>
>>
>> Ideally the function will return the extrema values and their
>> positions. Relatively simple function to implement in Python, but that
>> would be painfully slow. The typical data array I am looking at, has
>> approximately 500k elements of double precision float.
>
> I use the following to find the indices of extrema:
>
>
> def extrema_sampling(x, withend=False):
>    """
>    return the indices into the local maxima or minima of x
>    if withend, include the 0 and end points in the sampling
>    """
>    d = np.diff(x)
>    d2 = d[:-1]*d[1:]
>    ind = []
>    if withend:
>        ind.append(0)
>    ind.extend(np.nonzero(d2<0)[0]+1)
>    if withend and ind[-1]!=len(x)-1:
>        ind.append(len(x)-1)
>    return np.array(ind)

This is an excellent way of indexing extrema! Although it doesnt
return exactly what I wanted and it wont catch gradient changes to
zero.

Inspired by the above, I have come up with the following, hope it can
be of some use to others:


def extrema(x, max = True, min = True, strict = False, withend = False):
	"""
	This function will index the extrema of a given array x.
	
	Options:
		max		If true, will index maxima
		min		If true, will index minima
		strict		If true, will not index changes to zero gradient
		withend	If true, always include x[0] and x[-1]
	
	This function will return a tuple of extrema indexies and values
	"""
	
	# This is the gradient
	from numpy import zeros
	dx = zeros(len(x))
	from numpy import diff
	dx[1:] = diff(x)
	dx[0] = dx[1]
	
	# Clean up the gradient in order to pick out any change of sign
	from numpy import sign
	dx = sign(dx)
	
	# define the threshold for whether to pick out changes to zero gradient
	threshold = 0
	if strict:
		threshold = 1
		
	# Second order diff to pick out the spikes
	d2x = diff(dx)
	
	if max and min:
		d2x = abs(d2x)
	elif max:
		d2x = -d2x
	
	# Take care of the two ends
	if withend:
		d2x[0] = 2
		d2x[-1] = 2
	
	# Sift out the list of extremas
	from numpy import nonzero
	ind = nonzero(d2x > threshold)[0]
	
	return ind, x[ind]


From philbinj at gmail.com  Thu Oct  9 12:30:52 2008
From: philbinj at gmail.com (James Philbin)
Date: Thu, 9 Oct 2008 17:30:52 +0100
Subject: [SciPy-user] Left hand sparse matrix multiplication
In-Reply-To: <d05265cb0810081228r47731877ic187039ec71c0b6b@mail.gmail.com>
References: <2b1c8c4f0810080841l4df18c1fre07ee259a6164f74@mail.gmail.com>
	<d05265cb0810081228r47731877ic187039ec71c0b6b@mail.gmail.com>
Message-ID: <2b1c8c4f0810090930o36d96401o5f6b2b74f6fd2e8f@mail.gmail.com>

>> It seems there should be a much more efficient overload for csc's
>> rmatvec which doesn't involve computing the transpose. I hope i'm
>> understanding things correctly.
>>
>
> CSR.T and CSC.T are constant time operations, they just return the
> matrix in the "opposite" format.

Aha, great. It already does what I want.

Thanks,
James


From Dharhas.Pothina at twdb.state.tx.us  Thu Oct  9 13:02:10 2008
From: Dharhas.Pothina at twdb.state.tx.us (Dharhas Pothina)
Date: Thu, 09 Oct 2008 12:02:10 -0500
Subject: [SciPy-user] Calculating daily,
	monthly and	seasonal	averages	of hourly	time series data.
In-Reply-To: <1223563189.30896.47.camel@poste5>
References: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>
	<1223542287.26062.20.camel@poste5>
	<48EDB6AF.63BA.009B.0@twdb.state.tx.us>
	<1223563189.30896.47.camel@poste5>
Message-ID: <48EDF2C2.63BA.009B.0@twdb.state.tx.us>


This sounds great. I'm going to have to see how complicated it is to
get the recent versions of numpy/scipy/matplotlib installed on Fedora 8
so I can try the timeseries scikit out. From what I could tell there are
no repositories or rpms available for matplotlib 0.98 on Fedora 8 or 9.

thanks for your help.

- dharhas

>>> Lionel Roubeyrie <lroubeyrie at limair.asso.fr> 10/9/2008 9:39 AM >>>

> Ok I think I understood your example below. Can you give me an
example
> of how to deal with missing data?
If you take my last example, you have 3 separated arrays:
mes.data : your values
mes.dates : the date array
mes.mask : like the maskedarray module (timeseries is based on it).
Trying to mask the 10 first values:
##################################
In [16]: mask=zeros_like(mes.data)

In [17]: mask[0:10]=True

In [18]: mask
Out[18]: array([ 1.,  1.,  1., ...,  0.,  0.,  0.])

In [19]: mes2=ts.time_series(mes, mask=mask)

In [20]: mes2
Out[20]: 
timeseries([-- -- -- ..., 17.9699245692 66.8968405206 24.7117965045],
           dates = [01-jan-2007 00:00 ... 30-d?c-2008 23:00],
           freq  = H)
##################################
A timeseries can be constructed based on another timeseries, like I do
here with mes2. Note that just the values are masked (missing), not
the
dates because all fields have a value (masked or not).


>  Does this general technique work for data that is on a 15 minute
frequency 
Yes, but no :-) The timeseries module doesn't handle directly QH
frequency, but minute frequency (freq='T'). Look at that :
#################################
In [28]: fielddates=ts.date_array(['2007-01-01 00:00', '2007-01-01
00:15', '2007-01-01 00:30', '2007-01-01 00:45'], freq='T')

In [29]: salinity=random(4)*100

In [30]: mes=ts.time_series(data=salinity, dates=fielddates)

In [31]: mes.has_missing_dates()
Out[31]: True
#################################
There's not QH native frequency, then there's some missing dates (you
can also look for duplicated dates, very convenient!). But you can
fill
these missing dates :
###############################
In [36]: mes2=mes.fill_missing_dates()

In [37]: mes2.has_missing_dates()
Out[37]: False

In [38]: mes2
Out[38]: 
timeseries([2.33824586442 -- -- -- -- -- -- -- -- -- -- -- -- -- --
36.180901427 -- --
 -- -- -- -- -- -- -- -- -- -- -- -- 39.0648471531 -- -- -- -- -- --
--
--
 -- -- -- -- -- -- 55.4226606997],
           dates = [01-jan-2007 00:00 ... 01-jan-2007 00:45],
           freq  = T)
###############################
Or the module can handle directly these missing dates when you convert
the timeseries to a lower frequency:
###############################
In [39]: mes.convert(freq='H', func=mean)
Out[39]: 
timeseries([ 33.25166379],
           dates = [01-jan-2007 00:00],
           freq  = H)
###############################
You can try with func=None, it will just fill the missing dates with
missing values :-p

> or datasets where the frequency is
> variable (ie some months we have 10 readings other months we may
have
> 30?
Like you see, just pass you datas with the corrects dates, and it
rocks,
but don't mix minute frequency with hour frequency!
Here I take 3 daily samples in january, and one in october :
#################################
In [41]: fielddates=ts.date_array(['2007-01-01', '2007-01-02',
'2007-01-03', '2007-10-15'], freq='D')

In [42]: salinity=random(4)*100

In [43]: mes=ts.time_series(data=salinity, dates=fielddates)

In [44]: mes
Out[44]: 
timeseries([ 59.63468614  38.60721076  64.52554805  66.17637291],
           dates = [01-jan-2007 02-jan-2007 03-jan-2007 15-oct-2007],
           freq  = D)


In [45]: mes.convert(freq='M', func=mean)
Out[45]: 
timeseries([54.2558149823 -- -- -- -- -- -- -- -- 66.1763729106],
           dates = [jan-2007 ... oct-2007],
           freq  = M)
###################################
Computing the monthly average goes fine, the module fill the missing
months by masked values.

> 
> Also how stable is the scikits.timeseries? Is it reasonably usable?
Yes, we use it intensively on large projects and Pierre G.M. has made
a
very good tool.
Cordialy

> 
> thanks,
> 
> - dharhas
> 
> >>> Lionel Roubeyrie <lroubeyrie at limair.asso.fr> 10/9/2008 3:51 AM
>>>
> Hi Dharhas,
> scikits.timeseries is perfect for what you want in a very useable
way
> :
> 
> ###############################
> In [29]: import scikits.timeseries as ts
> 
> In [30]: sdate=ts.Date('H', '2007-01-01 00:00')
> 
> In [31]: fielddates=ts.date_array(start_date=sdate, freq='H',
> length=365*24*2)
> 
> In [32]: salinity=random(365*24*2)*100
> 
> In [33]: mes=ts.time_series(data=salinity, dates=fielddates)
> 
> In [34]: mes
> Out[34]: 
> timeseries([ 23.84116045  49.51437251  89.29221711 ...,  37.00510947
> 41.12589836
>   78.65572656],
>            dates = [01-jan-2007 00:00 ... 30-d?c-2008 23:00],
>            freq  = H)
> 
> 
> In [35]: mes_avmonth=mes.convert(freq='M', func=mean)
> 
> In [36]: mes_avmonth
> Out[36]: 
> timeseries([ 49.29718906  50.64688937  49.88193999  48.97144253
> 49.5788259
>   50.41340038  50.15047009  51.70933261  50.5635153   51.15084406
>   51.15362514  51.51443468  49.17556599  49.26877667  50.21416724
>   49.37037657  51.00724033  49.43337134  49.60398056  50.24470761
>   50.62350109  51.15572702  51.37652011  49.24193747],
>            dates = [jan-2007 ... d?c-2008],
>            freq  = M)
> 
> 
> In [37]: mes_avyear=mes.convert(freq='Y', func=mean)
> 
> In [38]: mes_avyear
> Out[38]: 
> timeseries([ 50.41903159  50.06468157],
>            dates = [2007 2008],
>            freq  = A-DEC)
> 
> 
> In [39]: mes_avseason=mes[(mes.month>=5) & (mes.month<=9)].mean()
> 
> In [40]: mes_avseason
> Out[40]: 50,33380690600049
> ###############################
> 
> 
> Le mercredi 08 octobre 2008 ? 14:54 -0500, Dharhas Pothina a ?crit :
> > Hi,
> > 
> > I'm trying to analyze hourly salinity data. I was wondering if
there
> is a simple way of calculating daily, monthly and seasonal averages
of
> hourly time series data. 
> > 
> > So assuming I have two arrays that contain several years of hourly
> (or every 15min) salinity data: a datetime array called 'fielddates'
& a
> data array called 'salinity'
> > 
> > How would I go about getting the various averages. The seasonal
> averages would be say defined as May through September etc. 
> > 
> > I had a look at scikits.timeseries but it looks like it would
require
> upgrading numpy to install and there isn't enough high level
> documentation on how to use it for me to be confident in picking it
up
> in the time frame I'm looking at. I'm also not completely clear if
it
> can handle stuff that happens on a scale smaller than a day. If
anyone
> can point me to any usage examples for it that would be appreciated.
> > 
> > Thanks,
> > 
> > - dharhas
> > 
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org 
> > http://projects.scipy.org/mailman/listinfo/scipy-user 
> > 
-- 
Lionel Roubeyrie
charg? d'?tudes
LIMAIR - La Surveillance de l'Air en Limousin
http://www.limair.asso.fr 


_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org 
http://projects.scipy.org/mailman/listinfo/scipy-user


From pgmdevlist at gmail.com  Thu Oct  9 13:17:28 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Thu, 9 Oct 2008 13:17:28 -0400
Subject: [SciPy-user]
 =?iso-8859-15?q?Calculating_daily=2C_monthly_and=09s?=
 =?iso-8859-15?q?easonal=09averages=09of_hourly=09time_series_data=2E?=
In-Reply-To: <48EDF2C2.63BA.009B.0@twdb.state.tx.us>
References: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>
	<1223563189.30896.47.camel@poste5>
	<48EDF2C2.63BA.009B.0@twdb.state.tx.us>
Message-ID: <200810091317.29158.pgmdevlist@gmail.com>

Dharhas,
What you need for timeseries is a recent numpy (>=1.2.0) and 
scipy(>=0.7svn...). If you can compile the latest sources from SVN, you're 
good to go.

The documentation is here:
http://pytseries.sourceforge.net/

If it doesn't cover areas you need, let us know and help us by writing some 
examples/tutorial/whatever.

As an extra comment on Lionel's answers:
* Once you have a time series, you can directly mask some data without having 
to recreate a series
>>> series=ts.time_series(np.random.rand(365), start_date=ts.now('D'))
>>># Mask the first 10 elements
>>> series[:10] = ma.masked
>>> #Maske the data in August
>>> series[series.month==8] = ma.masked


From nwagner at iam.uni-stuttgart.de  Thu Oct  9 13:35:32 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 09 Oct 2008 19:35:32 +0200
Subject: [SciPy-user] Wrapping mebdfdae.f
Message-ID: <web-112569104@uni-stuttgart.de>

Hi all,

I tried to wrap mebdfdae.f
http://www.ma.ic.ac.uk/~jcash/IVP_software_BSD/mebdfdae.f

If failed with

f2py -c -m mebdfdae mebdfdae.f
running build
running config_cc
unifing config_cc, config, build_clib, build_ext, build 
commands --compiler options
running config_fc
unifing config_fc, config, build_clib, build_ext, build 
commands --fcompiler options
running build_src
building extension "mebdfdae" sources
f2py options: []
f2py:> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c
creating /tmp/tmp9spUor
creating /tmp/tmp9spUor/src.linux-i686-2.4
Reading fortran codes...
         Reading file 'mebdfdae.f' (format:fix,strict)
Post-processing...
         Block: mebdfdae
                         Block: mebdf
                         Block: ovdriv
                         Block: interp
                         Block: coset
                         Block: pset
                                         Block: pderv
                                         Block: f
                         Block: dec
                         Block: sol
                         Block: dgbfa
                         Block: daxpy
                         Block: dscal
                         Block: idamax
                         Block: dgbsl
                         Block: ddot
                         Block: errors
                         Block: prdict
                                         Block: f
                         Block: itrat2
                                         Block: f
                         Block: stiff
                                         Block: f
                                         Block: mas
                         Block: rscale
                         Block: cpyary
                         Block: hchose
                         Block: dlamch
                         Block: dlamc1
                         Block: dlamc2
                         Block: dlamc3
                         Block: dlamc4
                         Block: dlamc5
                         Block: lsame
Post-processing (stage 2)...
Building modules...
         Constructing call-back function 
"cb_f_in_pset__user__routines"
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
           def f(t,y,wrkspc,ipar,rpar,ierr,[n]): return
         Constructing call-back function 
"cb_pderv_in_pset__user__routines"
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
           def pderv(t,y,pwcopy,ipar,rpar,ierr,[n,n]): 
return
         Constructing call-back function 
"cb_f_in_prdict__user__routines"
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
           def f(t,y,yprime,ipar,rpar,ierr,[n]): return
         Constructing call-back function 
"cb_f_in_itrat2__user__routines"
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
           def f(t,save1,save2,ipar,rpar,ierr,[n]): return
         Constructing call-back function 
"cb_f_in_stiff__user__routines"
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
           def f(t,y,save1,ipar,rpar,ierr,[n]): return
         Constructing call-back function 
"cb_mas_in_stiff__user__routines"
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
           def mas(n,am,ldmas,ipar,rpar,ierr): return
         Building module "mebdfdae"...
                 Constructing wrapper function "mebdf"...
routsign2map: Confused: function mebdf has externals ['f', 
'pderv', 'mas'] but no "use" statement.
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
sign2map: Confused: external f is not in lcb_map[].
append_needs: unknown need 'f'
append_needs: unknown need 'f'
sign2map: Confused: external pderv is not in lcb_map[].
append_needs: unknown need 'pderv'
append_needs: unknown need 'pderv'
sign2map: Confused: external mas is not in lcb_map[].
append_needs: unknown need 'mas'
append_needs: unknown need 'mas'
                   mebdf(t0,ho,y0,tout,tend,mf,idid,lout,work,iwork,mbnd,masbnd,maxder,itol,rtol,atol,rpar,ipar,f,pderv,mas,ierr,[n,lwork,liwork,f_extra_args,pderv_extra_args,mas_extra_args])
                 Constructing wrapper function "ovdriv"...
routsign2map: Confused: function ovdriv has externals 
['f', 'pderv', 'mas'] but no "use" statement.
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
sign2map: Confused: external f is not in lcb_map[].
append_needs: unknown need 'f'
append_needs: unknown need 'f'
sign2map: Confused: external pderv is not in lcb_map[].
append_needs: unknown need 'pderv'
append_needs: unknown need 'pderv'
sign2map: Confused: external mas is not in lcb_map[].
append_needs: unknown need 'mas'
append_needs: unknown need 'mas'
                   ovdriv(t0,ho,y0,tout,tend,mf,idid,lout,y,yhold,ynhold,ymax,errors,save1,save2,scale,arh,pw,pwcopy,am,ipiv,mbnd,masbnd,nind1,nind2,nind3,maxder,itol,rtol,atol,rpar,ipar,f,pderv,mas,nqused,nstep,nfail,nfe,nje,ndec,nbsol,npset,ncoset,maxord,maxstp,uround,hused,epsjac,ierr,[n,f_extra_args,pderv_extra_args,mas_extra_args])
                 Constructing wrapper function "interp"...
                   interp(jstart,h,t,y,tout,y0,[n])
                 Constructing wrapper function "coset"...
                   coset(nq,el,elst,tq,ncoset,maxord)
                 Constructing wrapper function "pset"...
sign2map: Confused: external mas is not in 
lcb_map['pderv', 'f'].
append_needs: unknown need 'mas'
append_needs: unknown need 'mas'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
                   pset(y,h,t,uround,epsjac,con,miter,mbnd,masbnd,nind1,nind2,nind3,ier,f,pderv,mas,nrenew,ymax,save1,save2,pw,pwcopy,am,wrkspc,ipiv,itol,rtol,atol,npset,nje,nfe,ndec,ipar,rpar,ierr,[n,f_extra_args,pderv_extra_args,mas_extra_args])
                 Constructing wrapper function "dec"...
                   dec(a,ip,ier,[n,ndim])
                 Constructing wrapper function "sol"...
                   sol(a,b,ip,[n,ndim])
                 Constructing wrapper function "dgbfa"...
                   dgbfa(abd,n,ml,mu,ipvt,info,[lda])
                 Constructing wrapper function "daxpy"...
                   daxpy(n,da,dx,incx,dy,incy)
                 Constructing wrapper function "dscal"...
                   dscal(n,da,dx,incx)
                 Creating wrapper for Fortran function 
"idamax"("idamax")...
                 Constructing wrapper function "idamax"...
                   idamax = idamax(n,dx,incx)
                 Constructing wrapper function "dgbsl"...
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
                   dgbsl(abd,n,ml,mu,ipvt,b,job,[lda])
                 Creating wrapper for Fortran function 
"ddot"("ddot")...
                 Constructing wrapper function "ddot"...
                   ddot = ddot(n,dx,incx,dy,incy)
                 Constructing wrapper function "errors"...
                   errors(n,tq,edn,e,eup,bnd,eddn)
                 Constructing wrapper function "prdict"...
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
                   prdict(t,h,y,l,yprime,nfe,ipar,rpar,f,ierr,[n,f_extra_args])
                 Constructing wrapper function "itrat2"...
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
                   itrat2(qqq,y,t,hbeta,errbnd,arh,crate,tcrate,m,worked,ymax,error,save1,save2,scale,pw,mf,mbnd,am,masbnd,nind1,nind2,nind3,ipiv,lmb,itol,rtol,atol,ipar,rpar,hused,nbsol,nfe,nqused,f,ierr,[n,f_extra_args])
                 Constructing wrapper function "stiff"...
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
getarrdims:warning: assumed shape array, using 0 instead 
of '*'
sign2map: Confused: external pderv is not in 
lcb_map['mas', 'f'].
append_needs: unknown need 'pderv'
append_needs: unknown need 'pderv'
                   stiff(h,hmax,hmin,jstart,kflag,mf,mbnd,masbnd,nind1,nind2,nind3,t,tout,tend,y,ymax,error,save1,save2,scale,pw,pwcopy,am,yhold,ynhold,arh,ipiv,lout,maxder,itol,rtol,atol,rpar,ipar,f,pderv,mas,nqused,nstep,nfail,nfe,nje,ndec,nbsol,npset,ncoset,maxord,maxstp,uround,epsjac,hused,ierr,[n,f_extra_args,pderv_extra_args,mas_extra_args])
                 Constructing wrapper function "rscale"...
                   rscale(l,rh,y,[n])
                 Constructing wrapper function "cpyary"...
                   cpyary(source,target,[nelem])
                 Constructing wrapper function "hchose"...
                   hchose(rh,h,ovride)
                 Creating wrapper for Fortran function 
"dlamch"("dlamch")...
                 Constructing wrapper function "dlamch"...
                   dlamch = dlamch(cmach)
                 Constructing wrapper function "dlamc1"...
                   dlamc1(beta,t,rnd,ieee1)
                 Constructing wrapper function "dlamc2"...
                   dlamc2(beta,t,rnd,eps,emin,rmin,emax,rmax)
                 Creating wrapper for Fortran function 
"dlamc3"("dlamc3")...
                 Constructing wrapper function "dlamc3"...
                   dlamc3 = dlamc3(a,b)
                 Constructing wrapper function "dlamc4"...
                   dlamc4(emin,start,base)
                 Constructing wrapper function "dlamc5"...
                   dlamc5(beta,p,emin,ieee,emax,rmax)
                 Creating wrapper for Fortran function 
"lsame"("lsame")...
                 Constructing wrapper function "lsame"...
                   lsame = lsame(ca,cb)
                 Constructing COMMON block support for 
"stpsze"...
                   hstpsz
         Wrote C/API module "mebdfdae" to file 
"/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c"
         Fortran 77 wrappers are saved to 
"/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdae-f2pywrappers.f"
   adding 
'/tmp/tmp9spUor/src.linux-i686-2.4/fortranobject.c' to 
sources.
   adding '/tmp/tmp9spUor/src.linux-i686-2.4' to 
include_dirs.
copying 
/usr/lib/python2.4/site-packages/numpy/f2py/src/fortranobject.c 
-> /tmp/tmp9spUor/src.linux-i686-2.4
copying 
/usr/lib/python2.4/site-packages/numpy/f2py/src/fortranobject.h 
-> /tmp/tmp9spUor/src.linux-i686-2.4
   adding 
'/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdae-f2pywrappers.f' 
to sources.
running build_ext
customize UnixCCompiler
customize UnixCCompiler using build_ext
customize GnuFCompiler
Found executable /usr/bin/g77
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler
gnu: no Fortran 90 compiler found
gnu: no Fortran 90 compiler found
customize GnuFCompiler using build_ext
building 'mebdfdae' extension
compiling C sources
C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -O2 
-march=i586 -mcpu=i686 -fmessage-length=0 -Wall -g -fPIC

creating /tmp/tmp9spUor/tmp
creating /tmp/tmp9spUor/tmp/tmp9spUor
creating /tmp/tmp9spUor/tmp/tmp9spUor/src.linux-i686-2.4
compile options: '-I/tmp/tmp9spUor/src.linux-i686-2.4 
-I/usr/lib/python2.4/site-packages/numpy/core/include 
-I/usr/include/python2.4 -c'
gcc: /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: redefinition of `n_cb_capi'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: `n_cb_capi' previously declared here
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `cb_pderv_in_pset__user__routines':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: redeclaration of `n_cb_capi'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: `n_cb_capi' previously declared here
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:608: 
error: redeclaration of `n'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607: 
error: `n' previously declared here
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607: 
warning: unused variable `n'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_mebdf':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: `f_typedef' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: (Each undeclared identifier is reported only once
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: for each function it appears in.)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: syntax error before "f_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472: 
error: `pderv_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472: 
error: syntax error before "pderv_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477: 
error: `mas_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477: 
error: syntax error before "mas_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1555: 
error: `pderv_cptr' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1557: 
error: `pderv' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1560: 
error: `pderv_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: `nofoptargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1572: 
error: `mas_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1574: 
error: `mas' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1577: 
error: `mas_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1612: 
error: `f_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1614: 
error: `f' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1617: 
error: `f_nofargs' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_ovdriv':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020: 
error: `f_typedef' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020: 
error: syntax error before "f_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025: 
error: `pderv_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025: 
error: syntax error before "pderv_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030: 
error: `mas_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030: 
error: syntax error before "mas_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2164: 
error: `pderv_cptr' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2166: 
error: `pderv' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2169: 
error: `pderv_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: `nofoptargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2184: 
error: `mas_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2186: 
error: `mas' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2189: 
error: `mas_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2201: 
error: `f_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2203: 
error: `f' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2206: 
error: `f_nofargs' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_pset':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027: 
error: `mas_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027: 
error: syntax error before "mas_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3228: 
error: `mas_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3230: 
error: `mas' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3233: 
error: `mas_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: `nofoptargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_stiff':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829: 
error: `pderv_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829: 
error: syntax error before "pderv_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5920: 
error: `pderv_cptr' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5922: 
error: `pderv' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5925: 
error: `pderv_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: `nofoptargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: redefinition of `n_cb_capi'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: `n_cb_capi' previously declared here
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `cb_pderv_in_pset__user__routines':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: redeclaration of `n_cb_capi'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599: 
error: `n_cb_capi' previously declared here
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:608: 
error: redeclaration of `n'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607: 
error: `n' previously declared here
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607: 
warning: unused variable `n'
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_mebdf':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: `f_typedef' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: (Each undeclared identifier is reported only once
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: for each function it appears in.)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467: 
error: syntax error before "f_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472: 
error: `pderv_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472: 
error: syntax error before "pderv_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477: 
error: `mas_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477: 
error: syntax error before "mas_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1555: 
error: `pderv_cptr' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1557: 
error: `pderv' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1560: 
error: `pderv_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561: 
error: `nofoptargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1572: 
error: `mas_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1574: 
error: `mas' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1577: 
error: `mas_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1612: 
error: `f_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1614: 
error: `f' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1617: 
error: `f_nofargs' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_ovdriv':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020: 
error: `f_typedef' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020: 
error: syntax error before "f_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025: 
error: `pderv_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025: 
error: syntax error before "pderv_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030: 
error: `mas_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030: 
error: syntax error before "mas_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2164: 
error: `pderv_cptr' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2166: 
error: `pderv' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2169: 
error: `pderv_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170: 
error: `nofoptargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2184: 
error: `mas_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2186: 
error: `mas' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2189: 
error: `mas_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2201: 
error: `f_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2203: 
error: `f' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2206: 
error: `f_nofargs' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_pset':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027: 
error: `mas_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027: 
error: syntax error before "mas_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3228: 
error: `mas_cptr' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3230: 
error: `mas' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3233: 
error: `mas_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234: 
error: `nofoptargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In 
function `f2py_rout_mebdfdae_stiff':
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829: 
error: `pderv_typedef' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829: 
error: syntax error before "pderv_cptr"
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5920: 
error: `pderv_cptr' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5922: 
error: `pderv' undeclared (first use in this function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5925: 
error: `pderv_nofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: `maxnofargs' undeclared (first use in this 
function)
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: syntax error at '#' token
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926: 
error: `nofoptargs' undeclared (first use in this 
function)
error: Command "gcc -pthread -fno-strict-aliasing -DNDEBUG 
-O2 -march=i586 -mcpu=i686 -fmessage-length=0 -Wall -g 
-fPIC -I/tmp/tmp9spUor/src.linux-i686-2.4 
-I/usr/lib/python2.4/site-packages/numpy/core/include 
-I/usr/include/python2.4 -c 
/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c -o 
/tmp/tmp9spUor/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.o" 
failed with exit status 1

How can I fix this problem ?

Nils


From matthew.brett at gmail.com  Thu Oct  9 14:13:03 2008
From: matthew.brett at gmail.com (Matthew Brett)
Date: Thu, 9 Oct 2008 11:13:03 -0700
Subject: [SciPy-user] Reference to algorithm for matrix rank
Message-ID: <1e2af89e0810091113o6af11cb7ue23a792e85be7f42@mail.gmail.com>

Hi,

I wanted to write a generic matrix rank algorithm.  The general form
seems to be standard:

def matrix_rank(M, tol):
   S = svd(M, compute_uv=False)
   return np.sum(S > tol)

but what I can't find is some citable reference for a general way to
choose 'tol'.  Does anyone know of the right source for this?

Thanks a lot,

Matthew


From bblais at bryant.edu  Thu Oct  9 14:26:50 2008
From: bblais at bryant.edu (Brian Blais)
Date: Thu, 9 Oct 2008 14:26:50 -0400
Subject: [SciPy-user] why does linalg.sqrtm return an array?
Message-ID: <7F046513-107E-4F6D-95F3-A160CF26A800@bryant.edu>

Hello,

I was just bitten by the fact that sqrtm returns an array, not a  
matrix.  Every time I tried to test it with Q*Q, or Q.T*Q I got very  
strange results (I was just about to post that it was seriously  
broken).  Then I found out that it returns an array!  Why is that?   
It would seem that if you are doing a matrix square root, you are  
working with matrices mostly in that calculation, so a matrix is the  
consistent thing to return (and perhaps the least surprising thing to  
return).  Is there a reason for this?


		thanks,

			Brian Blais

-- 
Brian Blais
bblais at bryant.edu
http://web.bryant.edu/~bblais



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081009/fa8b6510/attachment.html>

From nwagner at iam.uni-stuttgart.de  Thu Oct  9 14:30:40 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 09 Oct 2008 20:30:40 +0200
Subject: [SciPy-user] why does linalg.sqrtm return an array?
In-Reply-To: <7F046513-107E-4F6D-95F3-A160CF26A800@bryant.edu>
References: <7F046513-107E-4F6D-95F3-A160CF26A800@bryant.edu>
Message-ID: <web-112570547@uni-stuttgart.de>

On Thu, 9 Oct 2008 14:26:50 -0400
  Brian Blais <bblais at bryant.edu> wrote:
> Hello,
> 
> I was just bitten by the fact that sqrtm returns an 
>array, not a  matrix.  Every time I tried to test it with 
>Q*Q, or Q.T*Q I got very  strange results (I was just 
>about to post that it was seriously  broken).  Then I 
>found out that it returns an array!  Why is that?   It 
>would seem that if you are doing a matrix square root, 
>you are  working with matrices mostly in that 
>calculation, so a matrix is the  consistent thing to 
>return (and perhaps the least surprising thing to 
> return).  Is there a reason for this?
> 
> 
> 		thanks,
> 
> 			Brian Blais
> 
> -- 
> Brian Blais
> bblais at bryant.edu
> http://web.bryant.edu/~bblais
> 
> 
> 
This is a known problem.
See
http://projects.scipy.org/scipy/scipy/ticket/585

Nils


From Dharhas.Pothina at twdb.state.tx.us  Thu Oct  9 14:42:49 2008
From: Dharhas.Pothina at twdb.state.tx.us (Dharhas Pothina)
Date: Thu, 09 Oct 2008 13:42:49 -0500
Subject: [SciPy-user] Calculating daily,
	monthly	and	seasonal	averages	of hourly	time series data.
In-Reply-To: <200810091317.29158.pgmdevlist@gmail.com>
References: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>
	<1223563189.30896.47.camel@poste5>
	<48EDF2C2.63BA.009B.0@twdb.state.tx.us>
	<200810091317.29158.pgmdevlist@gmail.com>
Message-ID: <48EE0A59.63BA.009B.0@twdb.state.tx.us>

Pierre,

I did already find the documentation you linked to and it looks like a excellent reference for the individual functions and classes. What I felt was missing was an overview of how to use the main features of the package in the form of examples, tutorials etc.

I've never tried compiling the scipy sources from svn. If I can manage it, I'll try the timeseries toolkit out. Once I start using it, I'd be happy to document my experiences and see if I can come up with some examples/tutorials based on what I am doing.  

fyi the website says Numpy 1.2.1 or later.

- dharhas

>>> Pierre GM <pgmdevlist at gmail.com> 10/9/2008 12:17 PM >>>
Dharhas,
What you need for timeseries is a recent numpy (>=1.2.0) and 
scipy(>=0.7svn...). If you can compile the latest sources from SVN, you're 
good to go.

The documentation is here:
http://pytseries.sourceforge.net/ 

If it doesn't cover areas you need, let us know and help us by writing some 
examples/tutorial/whatever.

As an extra comment on Lionel's answers:
* Once you have a time series, you can directly mask some data without having 
to recreate a series
>>> series=ts.time_series(np.random.rand(365), start_date=ts.now('D'))
>>># Mask the first 10 elements
>>> series[:10] = ma.masked
>>> #Maske the data in August
>>> series[series.month==8] = ma.masked
_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org 
http://projects.scipy.org/mailman/listinfo/scipy-user



From bolme1234 at comcast.net  Thu Oct  9 15:06:33 2008
From: bolme1234 at comcast.net (David Bolme)
Date: Thu, 9 Oct 2008 13:06:33 -0600
Subject: [SciPy-user] Extrema finding
In-Reply-To: <bfb379e10810090920i290af59dtfdf2c1394ebc4b1f@mail.gmail.com>
References: <bfb379e10810090444i4e435eb8q5f5729005087881b@mail.gmail.com>
	<88e473830810090521qe247f41l9649f3c460e95d38@mail.gmail.com>
	<bfb379e10810090920i290af59dtfdf2c1394ebc4b1f@mail.gmail.com>
Message-ID: <F47B9FAE-1E8C-41D7-A658-C77DEE0B1536@comcast.net>

This is a version for finding extrema in a 2D array.  It requires the  
ndimage maximum/minimum filters.  For the one dimensional case  
substitute size=[3] or use maximum_filter1d.  I keep writing this code  
over and over.  I am surprised that there is not a general purpose  
extrema finding routine in scipy.

     def localMax(mat):
         mx = maximum_filter(mat, size=[3,3])
         mn = minimum_filter(mat, size=[3,3])
	
         # (mat == mx) true if pixel is equal to the local max
         # The next computation suppresses responses where
         # the function is flat.
         local_maxima = ((mat == mx) & (mat != mn))
	
         # Get the indices of the maxima
         extrema = nonzero(local_maxima)
         return extrema


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081009/049f0c03/attachment.html>

From Dharhas.Pothina at twdb.state.tx.us  Thu Oct  9 15:27:28 2008
From: Dharhas.Pothina at twdb.state.tx.us (Dharhas Pothina)
Date: Thu, 09 Oct 2008 14:27:28 -0500
Subject: [SciPy-user] numpy/scipy svn installation question.
Message-ID: <48EE14CF.63BA.009B.0@twdb.state.tx.us>

Hi,

I want to install the svn versions of numpy/scipy (and 0.98 version of matplotlib) so that I can install the scikits.timeseries package. I have an existing installation of older versions of numpy/scipy/matplotlib installed through the fedora package manager.

>From what I understand from the website I should be able to build numpy & scipy by using 

python setup.py install

Is there a way to install the svn versions alongside my existing installations. Assuming I can do that how do I choose which version to use in a script.

thanks

- dharhas






From robert.kern at gmail.com  Thu Oct  9 19:34:40 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 9 Oct 2008 18:34:40 -0500
Subject: [SciPy-user] why does linalg.sqrtm return an array?
In-Reply-To: <7F046513-107E-4F6D-95F3-A160CF26A800@bryant.edu>
References: <7F046513-107E-4F6D-95F3-A160CF26A800@bryant.edu>
Message-ID: <3d375d730810091634v3e751fb5k722735ded08a601c@mail.gmail.com>

On Thu, Oct 9, 2008 at 13:26, Brian Blais <bblais at bryant.edu> wrote:
> Hello,
> I was just bitten by the fact that sqrtm returns an array, not a matrix.
>  Every time I tried to test it with Q*Q, or Q.T*Q I got very strange results
> (I was just about to post that it was seriously broken).  Then I found out
> that it returns an array!  Why is that?  It would seem that if you are doing
> a matrix square root, you are working with matrices mostly in that
> calculation, so a matrix is the consistent thing to return (and perhaps the
> least surprising thing to return).  Is there a reason for this?

Arguably, it should return a matrix object if given a matrix object,
but it should never return a matrix object if given a pure ndarray.
Just because one is doing matrix operations doesn't mean one wants to
use matrix objects.

But the fact that it returns an ndarray when given a matrix object is
a bug, and you're welcome to fix it. It's not always easy to do,
though, so you will find many such functions which do not preserve the
type of the input(s).

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From robert.kern at gmail.com  Thu Oct  9 19:40:06 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 9 Oct 2008 18:40:06 -0500
Subject: [SciPy-user] Reference to algorithm for matrix rank
In-Reply-To: <1e2af89e0810091113o6af11cb7ue23a792e85be7f42@mail.gmail.com>
References: <1e2af89e0810091113o6af11cb7ue23a792e85be7f42@mail.gmail.com>
Message-ID: <3d375d730810091640i414d0725scd4374f0037b0499@mail.gmail.com>

On Thu, Oct 9, 2008 at 13:13, Matthew Brett <matthew.brett at gmail.com> wrote:
> Hi,
>
> I wanted to write a generic matrix rank algorithm.  The general form
> seems to be standard:
>
> def matrix_rank(M, tol):
>   S = svd(M, compute_uv=False)
>   return np.sum(S > tol)
>
> but what I can't find is some citable reference for a general way to
> choose 'tol'.  Does anyone know of the right source for this?

You should get the book _Matrix Computations_ by Golub and van Loan.
You actually want tol to be relative to S.max(), not an absolute
tolerance. I like this:

  np.sum(S > (S.max() * np.finfo(M.dtype).eps)

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From nwagner at iam.uni-stuttgart.de  Fri Oct 10 02:25:31 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 10 Oct 2008 08:25:31 +0200
Subject: [SciPy-user] Trouble with f2py
Message-ID: <web-112589946@uni-stuttgart.de>

Hi all,

If I run f2py I get

f2py
Traceback (most recent call last):
   File "/data/home/nwagner/local/bin/f2py", line 20, in ?
     from numpy.f2py import main
   File 
"/data/home/nwagner/local/lib/python2.5/site-packages/numpy/__init__.py", 
line 125, in ?
     import add_newdocs
   File 
"/data/home/nwagner/local/lib/python2.5/site-packages/numpy/add_newdocs.py", 
line 9, in ?
     from lib import add_newdoc
   File 
"/data/home/nwagner/local/lib/python2.5/site-packages/numpy/lib/__init__.py", 
line 4, in ?
     from type_check import *
   File 
"/data/home/nwagner/local/lib/python2.5/site-packages/numpy/lib/type_check.py", 
line 8, in ?
     import numpy.core.numeric as _nx
   File 
"/data/home/nwagner/local/lib/python2.5/site-packages/numpy/core/__init__.py", 
line 5, in ?
     import multiarray
ImportError: 
/data/home/nwagner/local/lib/python2.5/site-packages/numpy/core/multiarray.so: 
undefined symbol: PyUnicodeUCS2_FromUnicode
  
How can I resolve this problem ?

Nils


From robert.kern at gmail.com  Fri Oct 10 02:29:40 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 10 Oct 2008 01:29:40 -0500
Subject: [SciPy-user] Trouble with f2py
In-Reply-To: <web-112589946@uni-stuttgart.de>
References: <web-112589946@uni-stuttgart.de>
Message-ID: <3d375d730810092329nf441f9l483e6181fb02aeeb@mail.gmail.com>

On Fri, Oct 10, 2008 at 01:25, Nils Wagner <nwagner at iam.uni-stuttgart.de> wrote:
> Hi all,
>
> If I run f2py I get
>
> f2py
> Traceback (most recent call last):
>   File "/data/home/nwagner/local/bin/f2py", line 20, in ?
>     from numpy.f2py import main
>   File
> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/__init__.py",
> line 125, in ?
>     import add_newdocs
>   File
> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/add_newdocs.py",
> line 9, in ?
>     from lib import add_newdoc
>   File
> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/lib/__init__.py",
> line 4, in ?
>     from type_check import *
>   File
> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/lib/type_check.py",
> line 8, in ?
>     import numpy.core.numeric as _nx
>   File
> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/core/__init__.py",
> line 5, in ?
>     import multiarray
> ImportError:
> /data/home/nwagner/local/lib/python2.5/site-packages/numpy/core/multiarray.so:
> undefined symbol: PyUnicodeUCS2_FromUnicode
>
> How can I resolve this problem ?

It looks like you installed a numpy binary built with a Python built
with UCS2 Unicode support while your Python binary was built with UCS4
Unicode support. Where did you get your Python? Where did you get your
numpy?

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From nwagner at iam.uni-stuttgart.de  Fri Oct 10 02:46:37 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Fri, 10 Oct 2008 08:46:37 +0200
Subject: [SciPy-user] Trouble with f2py
In-Reply-To: <3d375d730810092329nf441f9l483e6181fb02aeeb@mail.gmail.com>
References: <web-112589946@uni-stuttgart.de>
	<3d375d730810092329nf441f9l483e6181fb02aeeb@mail.gmail.com>
Message-ID: <web-112591035@uni-stuttgart.de>

On Fri, 10 Oct 2008 01:29:40 -0500
  "Robert Kern" <robert.kern at gmail.com> wrote:
> On Fri, Oct 10, 2008 at 01:25, Nils Wagner 
><nwagner at iam.uni-stuttgart.de> wrote:
>> Hi all,
>>
>> If I run f2py I get
>>
>> f2py
>> Traceback (most recent call last):
>>   File "/data/home/nwagner/local/bin/f2py", line 20, in 
>>?
>>     from numpy.f2py import main
>>   File
>> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/__init__.py",
>> line 125, in ?
>>     import add_newdocs
>>   File
>> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/add_newdocs.py",
>> line 9, in ?
>>     from lib import add_newdoc
>>   File
>> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/lib/__init__.py",
>> line 4, in ?
>>     from type_check import *
>>   File
>> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/lib/type_check.py",
>> line 8, in ?
>>     import numpy.core.numeric as _nx
>>   File
>> "/data/home/nwagner/local/lib/python2.5/site-packages/numpy/core/__init__.py",
>> line 5, in ?
>>     import multiarray
>> ImportError:
>> /data/home/nwagner/local/lib/python2.5/site-packages/numpy/core/multiarray.so:
>> undefined symbol: PyUnicodeUCS2_FromUnicode
>>
>> How can I resolve this problem ?
> 
> It looks like you installed a numpy binary built with a 
>Python built
> with UCS2 Unicode support while your Python binary was 
>built with UCS4
> Unicode support. Where did you get your Python? Where 
>did you get your
> numpy?
> 
Hi Robert,

I have modified the first line in f2py from

#!/usr/bin/env python

to

#!/usr/bin/env /data/home/nwagner/local/bin/python

Now it works. Sorry for the noise.

BTW, can you reproduce the problem concerning mebdfdae.f 
and f2py (my previous mail on the list) ?

Nils
  


From washakie at gmail.com  Fri Oct 10 08:45:02 2008
From: washakie at gmail.com (John [H2O])
Date: Fri, 10 Oct 2008 05:45:02 -0700 (PDT)
Subject: [SciPy-user]  scipy sclicing
Message-ID: <19917625.post@talk.nabble.com>


Could someone explain what I'm doing wrong here?

>>> i = array(range(140,149))
>>> j = array(range(5,20))
>>> i
array([140, 141, 142, 143, 144, 145, 146, 147, 148])
>>> j
array([ 5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19])
>>> a = acc[i,j]
Traceback (most recent call last):
  File "<string>", line 1, in <string>
ValueError: shape mismatch: objects cannot be broadcast to a single shape
>>> a = acc[140:148,5:19]
>>> 


I'm just following the indexing arrays usage on this page:
http://pages.physics.cornell.edu/~myers/teaching/ComputationalMethods/python/arrays.html

How come I can't use arrays to index my array?

Thanks!
-- 
View this message in context: http://www.nabble.com/scipy-sclicing-tp19917625p19917625.html
Sent from the Scipy-User mailing list archive at Nabble.com.



From stefan at sun.ac.za  Fri Oct 10 09:28:24 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Fri, 10 Oct 2008 15:28:24 +0200
Subject: [SciPy-user] scipy sclicing
In-Reply-To: <19917625.post@talk.nabble.com>
References: <19917625.post@talk.nabble.com>
Message-ID: <9457e7c80810100628x246291e4o23f5ae0bfb7885ed@mail.gmail.com>

2008/10/10 John [H2O] <washakie at gmail.com>:
>
> Could someone explain what I'm doing wrong here?
>
>>>> i = array(range(140,149))
>>>> j = array(range(5,20))
>>>> i
> array([140, 141, 142, 143, 144, 145, 146, 147, 148])
>>>> j
> array([ 5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19])
>>>> a = acc[i,j]
> Traceback (most recent call last):
>  File "<string>", line 1, in <string>
> ValueError: shape mismatch: objects cannot be broadcast to a single shape

The shapes of your indices,

In [5]: a.shape
Out[5]: (9,)

In [6]: b.shape
Out[6]: (15,)

cannot be broadcast to a single shape.  Either give the same number of
indices in i and j, or use i[:,None] and j.

Cheers
St?fan


From pearu at cens.ioc.ee  Fri Oct 10 09:03:49 2008
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Fri, 10 Oct 2008 16:03:49 +0300 (EEST)
Subject: [SciPy-user] Wrapping mebdfdae.f
In-Reply-To: <web-112569104@uni-stuttgart.de>
References: <web-112569104@uni-stuttgart.de>
Message-ID: <59715.172.17.0.4.1223643829.squirrel@cens.ioc.ee>


Hi,

The problem is that some Fortran functions in mebdfdae.f
take external pderv argument that signature f2py cannot
determine automatically. If you really need to call such
Fortran functions then you need to create a pyf file
and define pderv signature there.
Otherwise, just wrap only those functions that
you need to access from Python:

  f2py -c -m mebdfdae mebdfdae.f only: <function names that you need to
call from python>

HTH,
Pearu

On Thu, October 9, 2008 8:35 pm, Nils Wagner wrote:
> Hi all,
>
> I tried to wrap mebdfdae.f
> http://www.ma.ic.ac.uk/~jcash/IVP_software_BSD/mebdfdae.f
>
> If failed with
>
> f2py -c -m mebdfdae mebdfdae.f
> running build
> running config_cc
> unifing config_cc, config, build_clib, build_ext, build
> commands --compiler options
> running config_fc
> unifing config_fc, config, build_clib, build_ext, build
> commands --fcompiler options
> running build_src
> building extension "mebdfdae" sources
> f2py options: []
> f2py:> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c
> creating /tmp/tmp9spUor
> creating /tmp/tmp9spUor/src.linux-i686-2.4
> Reading fortran codes...
>          Reading file 'mebdfdae.f' (format:fix,strict)
> Post-processing...
>          Block: mebdfdae
>                          Block: mebdf
>                          Block: ovdriv
>                          Block: interp
>                          Block: coset
>                          Block: pset
>                                          Block: pderv
>                                          Block: f
>                          Block: dec
>                          Block: sol
>                          Block: dgbfa
>                          Block: daxpy
>                          Block: dscal
>                          Block: idamax
>                          Block: dgbsl
>                          Block: ddot
>                          Block: errors
>                          Block: prdict
>                                          Block: f
>                          Block: itrat2
>                                          Block: f
>                          Block: stiff
>                                          Block: f
>                                          Block: mas
>                          Block: rscale
>                          Block: cpyary
>                          Block: hchose
>                          Block: dlamch
>                          Block: dlamc1
>                          Block: dlamc2
>                          Block: dlamc3
>                          Block: dlamc4
>                          Block: dlamc5
>                          Block: lsame
> Post-processing (stage 2)...
> Building modules...
>          Constructing call-back function
> "cb_f_in_pset__user__routines"
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>            def f(t,y,wrkspc,ipar,rpar,ierr,[n]): return
>          Constructing call-back function
> "cb_pderv_in_pset__user__routines"
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>            def pderv(t,y,pwcopy,ipar,rpar,ierr,[n,n]):
> return
>          Constructing call-back function
> "cb_f_in_prdict__user__routines"
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>            def f(t,y,yprime,ipar,rpar,ierr,[n]): return
>          Constructing call-back function
> "cb_f_in_itrat2__user__routines"
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>            def f(t,save1,save2,ipar,rpar,ierr,[n]): return
>          Constructing call-back function
> "cb_f_in_stiff__user__routines"
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>            def f(t,y,save1,ipar,rpar,ierr,[n]): return
>          Constructing call-back function
> "cb_mas_in_stiff__user__routines"
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>            def mas(n,am,ldmas,ipar,rpar,ierr): return
>          Building module "mebdfdae"...
>                  Constructing wrapper function "mebdf"...
> routsign2map: Confused: function mebdf has externals ['f',
> 'pderv', 'mas'] but no "use" statement.
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> sign2map: Confused: external f is not in lcb_map[].
> append_needs: unknown need 'f'
> append_needs: unknown need 'f'
> sign2map: Confused: external pderv is not in lcb_map[].
> append_needs: unknown need 'pderv'
> append_needs: unknown need 'pderv'
> sign2map: Confused: external mas is not in lcb_map[].
> append_needs: unknown need 'mas'
> append_needs: unknown need 'mas'
>                    mebdf(t0,ho,y0,tout,tend,mf,idid,lout,work,iwork,mbnd,masbnd,maxder,itol,rtol,atol,rpar,ipar,f,pderv,mas,ierr,[n,lwork,liwork,f_extra_args,pderv_extra_args,mas_extra_args])
>                  Constructing wrapper function "ovdriv"...
> routsign2map: Confused: function ovdriv has externals
> ['f', 'pderv', 'mas'] but no "use" statement.
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> sign2map: Confused: external f is not in lcb_map[].
> append_needs: unknown need 'f'
> append_needs: unknown need 'f'
> sign2map: Confused: external pderv is not in lcb_map[].
> append_needs: unknown need 'pderv'
> append_needs: unknown need 'pderv'
> sign2map: Confused: external mas is not in lcb_map[].
> append_needs: unknown need 'mas'
> append_needs: unknown need 'mas'
>                    ovdriv(t0,ho,y0,tout,tend,mf,idid,lout,y,yhold,ynhold,ymax,errors,save1,save2,scale,arh,pw,pwcopy,am,ipiv,mbnd,masbnd,nind1,nind2,nind3,maxder,itol,rtol,atol,rpar,ipar,f,pderv,mas,nqused,nstep,nfail,nfe,nje,ndec,nbsol,npset,ncoset,maxord,maxstp,uround,hused,epsjac,ierr,[n,f_extra_args,pderv_extra_args,mas_extra_args])
>                  Constructing wrapper function "interp"...
>                    interp(jstart,h,t,y,tout,y0,[n])
>                  Constructing wrapper function "coset"...
>                    coset(nq,el,elst,tq,ncoset,maxord)
>                  Constructing wrapper function "pset"...
> sign2map: Confused: external mas is not in
> lcb_map['pderv', 'f'].
> append_needs: unknown need 'mas'
> append_needs: unknown need 'mas'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>                    pset(y,h,t,uround,epsjac,con,miter,mbnd,masbnd,nind1,nind2,nind3,ier,f,pderv,mas,nrenew,ymax,save1,save2,pw,pwcopy,am,wrkspc,ipiv,itol,rtol,atol,npset,nje,nfe,ndec,ipar,rpar,ierr,[n,f_extra_args,pderv_extra_args,mas_extra_args])
>                  Constructing wrapper function "dec"...
>                    dec(a,ip,ier,[n,ndim])
>                  Constructing wrapper function "sol"...
>                    sol(a,b,ip,[n,ndim])
>                  Constructing wrapper function "dgbfa"...
>                    dgbfa(abd,n,ml,mu,ipvt,info,[lda])
>                  Constructing wrapper function "daxpy"...
>                    daxpy(n,da,dx,incx,dy,incy)
>                  Constructing wrapper function "dscal"...
>                    dscal(n,da,dx,incx)
>                  Creating wrapper for Fortran function
> "idamax"("idamax")...
>                  Constructing wrapper function "idamax"...
>                    idamax = idamax(n,dx,incx)
>                  Constructing wrapper function "dgbsl"...
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>                    dgbsl(abd,n,ml,mu,ipvt,b,job,[lda])
>                  Creating wrapper for Fortran function
> "ddot"("ddot")...
>                  Constructing wrapper function "ddot"...
>                    ddot = ddot(n,dx,incx,dy,incy)
>                  Constructing wrapper function "errors"...
>                    errors(n,tq,edn,e,eup,bnd,eddn)
>                  Constructing wrapper function "prdict"...
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>                    prdict(t,h,y,l,yprime,nfe,ipar,rpar,f,ierr,[n,f_extra_args])
>                  Constructing wrapper function "itrat2"...
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
>                    itrat2(qqq,y,t,hbeta,errbnd,arh,crate,tcrate,m,worked,ymax,error,save1,save2,scale,pw,mf,mbnd,am,masbnd,nind1,nind2,nind3,ipiv,lmb,itol,rtol,atol,ipar,rpar,hused,nbsol,nfe,nqused,f,ierr,[n,f_extra_args])
>                  Constructing wrapper function "stiff"...
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> getarrdims:warning: assumed shape array, using 0 instead
> of '*'
> sign2map: Confused: external pderv is not in
> lcb_map['mas', 'f'].
> append_needs: unknown need 'pderv'
> append_needs: unknown need 'pderv'
>                    stiff(h,hmax,hmin,jstart,kflag,mf,mbnd,masbnd,nind1,nind2,nind3,t,tout,tend,y,ymax,error,save1,save2,scale,pw,pwcopy,am,yhold,ynhold,arh,ipiv,lout,maxder,itol,rtol,atol,rpar,ipar,f,pderv,mas,nqused,nstep,nfail,nfe,nje,ndec,nbsol,npset,ncoset,maxord,maxstp,uround,epsjac,hused,ierr,[n,f_extra_args,pderv_extra_args,mas_extra_args])
>                  Constructing wrapper function "rscale"...
>                    rscale(l,rh,y,[n])
>                  Constructing wrapper function "cpyary"...
>                    cpyary(source,target,[nelem])
>                  Constructing wrapper function "hchose"...
>                    hchose(rh,h,ovride)
>                  Creating wrapper for Fortran function
> "dlamch"("dlamch")...
>                  Constructing wrapper function "dlamch"...
>                    dlamch = dlamch(cmach)
>                  Constructing wrapper function "dlamc1"...
>                    dlamc1(beta,t,rnd,ieee1)
>                  Constructing wrapper function "dlamc2"...
>                    dlamc2(beta,t,rnd,eps,emin,rmin,emax,rmax)
>                  Creating wrapper for Fortran function
> "dlamc3"("dlamc3")...
>                  Constructing wrapper function "dlamc3"...
>                    dlamc3 = dlamc3(a,b)
>                  Constructing wrapper function "dlamc4"...
>                    dlamc4(emin,start,base)
>                  Constructing wrapper function "dlamc5"...
>                    dlamc5(beta,p,emin,ieee,emax,rmax)
>                  Creating wrapper for Fortran function
> "lsame"("lsame")...
>                  Constructing wrapper function "lsame"...
>                    lsame = lsame(ca,cb)
>                  Constructing COMMON block support for
> "stpsze"...
>                    hstpsz
>          Wrote C/API module "mebdfdae" to file
> "/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c"
>          Fortran 77 wrappers are saved to
> "/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdae-f2pywrappers.f"
>    adding
> '/tmp/tmp9spUor/src.linux-i686-2.4/fortranobject.c' to
> sources.
>    adding '/tmp/tmp9spUor/src.linux-i686-2.4' to
> include_dirs.
> copying
> /usr/lib/python2.4/site-packages/numpy/f2py/src/fortranobject.c
> -> /tmp/tmp9spUor/src.linux-i686-2.4
> copying
> /usr/lib/python2.4/site-packages/numpy/f2py/src/fortranobject.h
> -> /tmp/tmp9spUor/src.linux-i686-2.4
>    adding
> '/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdae-f2pywrappers.f'
> to sources.
> running build_ext
> customize UnixCCompiler
> customize UnixCCompiler using build_ext
> customize GnuFCompiler
> Found executable /usr/bin/g77
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler
> gnu: no Fortran 90 compiler found
> gnu: no Fortran 90 compiler found
> customize GnuFCompiler using build_ext
> building 'mebdfdae' extension
> compiling C sources
> C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -O2
> -march=i586 -mcpu=i686 -fmessage-length=0 -Wall -g -fPIC
>
> creating /tmp/tmp9spUor/tmp
> creating /tmp/tmp9spUor/tmp/tmp9spUor
> creating /tmp/tmp9spUor/tmp/tmp9spUor/src.linux-i686-2.4
> compile options: '-I/tmp/tmp9spUor/src.linux-i686-2.4
> -I/usr/lib/python2.4/site-packages/numpy/core/include
> -I/usr/include/python2.4 -c'
> gcc: /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: redefinition of `n_cb_capi'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: `n_cb_capi' previously declared here
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `cb_pderv_in_pset__user__routines':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: redeclaration of `n_cb_capi'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: `n_cb_capi' previously declared here
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:608:
> error: redeclaration of `n'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607:
> error: `n' previously declared here
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607:
> warning: unused variable `n'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_mebdf':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: `f_typedef' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: (Each undeclared identifier is reported only once
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: for each function it appears in.)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: syntax error before "f_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472:
> error: `pderv_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472:
> error: syntax error before "pderv_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477:
> error: `mas_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477:
> error: syntax error before "mas_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1555:
> error: `pderv_cptr' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1557:
> error: `pderv' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1560:
> error: `pderv_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: `nofoptargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1572:
> error: `mas_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1574:
> error: `mas' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1577:
> error: `mas_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1612:
> error: `f_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1614:
> error: `f' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1617:
> error: `f_nofargs' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_ovdriv':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020:
> error: `f_typedef' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020:
> error: syntax error before "f_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025:
> error: `pderv_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025:
> error: syntax error before "pderv_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030:
> error: `mas_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030:
> error: syntax error before "mas_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2164:
> error: `pderv_cptr' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2166:
> error: `pderv' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2169:
> error: `pderv_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: `nofoptargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2184:
> error: `mas_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2186:
> error: `mas' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2189:
> error: `mas_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2201:
> error: `f_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2203:
> error: `f' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2206:
> error: `f_nofargs' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_pset':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027:
> error: `mas_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027:
> error: syntax error before "mas_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3228:
> error: `mas_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3230:
> error: `mas' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3233:
> error: `mas_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: `nofoptargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_stiff':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829:
> error: `pderv_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829:
> error: syntax error before "pderv_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5920:
> error: `pderv_cptr' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5922:
> error: `pderv' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5925:
> error: `pderv_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: `nofoptargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: redefinition of `n_cb_capi'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: `n_cb_capi' previously declared here
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `cb_pderv_in_pset__user__routines':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: redeclaration of `n_cb_capi'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:599:
> error: `n_cb_capi' previously declared here
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:608:
> error: redeclaration of `n'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607:
> error: `n' previously declared here
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:607:
> warning: unused variable `n'
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_mebdf':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: `f_typedef' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: (Each undeclared identifier is reported only once
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: for each function it appears in.)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1467:
> error: syntax error before "f_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472:
> error: `pderv_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1472:
> error: syntax error before "pderv_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477:
> error: `mas_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1477:
> error: syntax error before "mas_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1555:
> error: `pderv_cptr' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1557:
> error: `pderv' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1560:
> error: `pderv_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1561:
> error: `nofoptargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1572:
> error: `mas_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1574:
> error: `mas' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1577:
> error: `mas_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1578:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1612:
> error: `f_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1614:
> error: `f' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1617:
> error: `f_nofargs' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:1618:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_ovdriv':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020:
> error: `f_typedef' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2020:
> error: syntax error before "f_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025:
> error: `pderv_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2025:
> error: syntax error before "pderv_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030:
> error: `mas_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2030:
> error: syntax error before "mas_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2164:
> error: `pderv_cptr' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2166:
> error: `pderv' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2169:
> error: `pderv_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2170:
> error: `nofoptargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2184:
> error: `mas_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2186:
> error: `mas' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2189:
> error: `mas_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2190:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2201:
> error: `f_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2203:
> error: `f' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2206:
> error: `f_nofargs' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:2207:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_pset':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027:
> error: `mas_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3027:
> error: syntax error before "mas_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3228:
> error: `mas_cptr' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3230:
> error: `mas' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3233:
> error: `mas_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:3234:
> error: `nofoptargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c: In
> function `f2py_rout_mebdfdae_stiff':
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829:
> error: `pderv_typedef' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5829:
> error: syntax error before "pderv_cptr"
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5920:
> error: `pderv_cptr' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5922:
> error: `pderv' undeclared (first use in this function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5925:
> error: `pderv_nofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: `maxnofargs' undeclared (first use in this
> function)
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: syntax error at '#' token
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c:5926:
> error: `nofoptargs' undeclared (first use in this
> function)
> error: Command "gcc -pthread -fno-strict-aliasing -DNDEBUG
> -O2 -march=i586 -mcpu=i686 -fmessage-length=0 -Wall -g
> -fPIC -I/tmp/tmp9spUor/src.linux-i686-2.4
> -I/usr/lib/python2.4/site-packages/numpy/core/include
> -I/usr/include/python2.4 -c
> /tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.c -o
> /tmp/tmp9spUor/tmp/tmp9spUor/src.linux-i686-2.4/mebdfdaemodule.o"
> failed with exit status 1
>
> How can I fix this problem ?
>
> Nils
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>




From washakie at gmail.com  Fri Oct 10 09:49:07 2008
From: washakie at gmail.com (John [H2O])
Date: Fri, 10 Oct 2008 06:49:07 -0700 (PDT)
Subject: [SciPy-user] scipy sclicing
In-Reply-To: <19917625.post@talk.nabble.com>
References: <19917625.post@talk.nabble.com>
Message-ID: <19918233.post@talk.nabble.com>


This seems to work:

def slize(X,i,j):
    X = X[i,:]
    X = X[:,j]
    return X

Problems with the approach?

-- 
View this message in context: http://www.nabble.com/scipy-sclicing-tp19917625p19918233.html
Sent from the Scipy-User mailing list archive at Nabble.com.



From pav at iki.fi  Fri Oct 10 09:58:40 2008
From: pav at iki.fi (Pauli Virtanen)
Date: Fri, 10 Oct 2008 13:58:40 +0000 (UTC)
Subject: [SciPy-user] scipy sclicing
References: <19917625.post@talk.nabble.com>
Message-ID: <gcnn2f$2ag$1@ger.gmane.org>

Fri, 10 Oct 2008 05:45:02 -0700, John [H2O] wrote:
> Could someone explain what I'm doing wrong here?
> 
>>>> i = array(range(140,149))
>>>> j = array(range(5,20))
>>>> i
> array([140, 141, 142, 143, 144, 145, 146, 147, 148])
>>>> j
> array([ 5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19])
>>>> a = acc[i,j]
> Traceback (most recent call last):
>   File "<string>", line 1, in <string>
> ValueError: shape mismatch: objects cannot be broadcast to a single
> shape
[clip]
> How come I can't use arrays to index my array?

See also the reference documentation; chapter 3.4 in Guide to Numpy [1].

.. [1] http://www.tramy.us/numpybook.pdf

-- 
Pauli Virtanen



From peridot.faceted at gmail.com  Fri Oct 10 09:59:21 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Fri, 10 Oct 2008 09:59:21 -0400
Subject: [SciPy-user] scipy sclicing
In-Reply-To: <19918233.post@talk.nabble.com>
References: <19917625.post@talk.nabble.com> <19918233.post@talk.nabble.com>
Message-ID: <ce557a360810100659m23e3c8b9p6a6f5b3011d3bd24@mail.gmail.com>

2008/10/10 John [H2O] <washakie at gmail.com>:
>
> This seems to work:
>
> def slize(X,i,j):
>    X = X[i,:]
>    X = X[:,j]
>    return X
>
> Problems with the approach?

No problem, per se, but it's a bit inefficient. I think the problem is
that multidimensional slicing with arrays doesn't work quite the way
you think it does. Let's say I have a 10 by 10 array X and I want rows
 1, 3, and 5, and columns 2 and 4. I can't write

X[ np.array([1,3,5]), np.array([2,4]) ]

because that's not how numpy's fancy indexing works. When you supply
arrays of indices like this (as opposed to slices), the idea is that
you're picking out arbitrary collections of elements, not just
rectangular hunks. For example if I want elements (1,2), (3,4), and
(5,0) I can write:

X[ np.array([1,3,5]), np.array([2,4,0]) ]

But what if you want a rectangular slice? Naively, you would have to
construct two big arrays:

X[ np.array([[1,1],[3,3],[5,5]]), np.array([[2,4],[2,4],[2,4]]) ]

This will give you what you want, but building those index arrays is
annoying. Fortunately numpy's broadcasting can do it for you,
repeating each array along a new axis:

X[ np.array([1,3,5])[:,np.newaxis], np.array([2,4])[np.newaxis,:] ]


Anne


From pgmdevlist at gmail.com  Fri Oct 10 11:21:50 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Fri, 10 Oct 2008 11:21:50 -0400
Subject: [SciPy-user]
 =?iso-8859-1?q?Calculating_daily=2C_monthly=09and=09?=
 =?iso-8859-1?q?seasonal=09averages=09of_hourly=09time_series_data=2E?=
In-Reply-To: <48EE0A59.63BA.009B.0@twdb.state.tx.us>
References: <48ECC9A6.63BA.009B.0@twdb.state.tx.us>
	<200810091317.29158.pgmdevlist@gmail.com>
	<48EE0A59.63BA.009B.0@twdb.state.tx.us>
Message-ID: <200810101121.50966.pgmdevlist@gmail.com>

Dharhas,

I agree that the documentation might be lacking a tutorial. Feel free to start 
one and we will add it on the site.

> fyi the website says Numpy 1.2.1 or later.

1.2.1 will have some bug fixes on numpy.ma that timeseries should rely on.


From philbinj at gmail.com  Fri Oct 10 11:34:40 2008
From: philbinj at gmail.com (James Philbin)
Date: Fri, 10 Oct 2008 16:34:40 +0100
Subject: [SciPy-user] Sparse sub-matrix indexing
Message-ID: <2b1c8c4f0810100834w2f2023b0jd78241d70bae377c@mail.gmail.com>

Hi,

I've come across this inconsistency with indexing sparse matrices:
In [3]: A = np.arange(100).reshape(10,10)
In [4]: B = spsp.csr_matrix(A)
In [5]: A[np.array([2]),:][:,np.array([2])]
Out[5]: array([[22]])
In [6]: B[np.array([2]),:][:,np.array([2])]
---------------------------------------------------------------------------
<type 'exceptions.TypeError'>             Traceback (most recent call last)

/home/james/<ipython console> in <module>()

/usr/lib/python2.5/site-packages/scipy/sparse/csr.py in __getitem__(self, key)
    226                 #[1:2,??]
    227                 if isintlike(col) or isinstance(col, slice):
--> 228                     return self._get_submatrix(row, col)      #[1:2,j]
    229                 else:
    230                     P = extractor(col,self.shape[1]).T
#[1:2,[1,2]]

/usr/lib/python2.5/site-packages/scipy/sparse/csr.py in
_get_submatrix(self, row_slice, col_slice)
    355
    356         i0, i1 = process_slice( row_slice, M )
--> 357         j0, j1 = process_slice( col_slice, N )
    358         check_bounds( i0, i1, M )
    359         check_bounds( j0, j1, N )

/usr/lib/python2.5/site-packages/scipy/sparse/csr.py in process_slice(sl, num)
    346
    347             else:
--> 348                 raise TypeError('expected slice or scalar')
    349
    350         def check_bounds( i0, i1, num ):

<type 'exceptions.TypeError'>: expected slice or scalar

However, this works fine:
In [7]: B[np.array([2,3]),:][:,np.array([2,3])]
Out[7]:
<2x2 sparse matrix of type '<type 'numpy.float64'>'
	with 4 stored elements in Compressed Sparse Row format>

Also, this works fine:
In [8]: B[[2],:][:,[2]]
Out[8]:
<1x1 sparse matrix of type '<type 'numpy.float64'>'
	with 1 stored elements in Compressed Sparse Row format>

Thanks,
James


From philbinj at gmail.com  Fri Oct 10 12:17:51 2008
From: philbinj at gmail.com (James Philbin)
Date: Fri, 10 Oct 2008 17:17:51 +0100
Subject: [SciPy-user] scipy.sparse.linalg.eigen inconsistency
Message-ID: <2b1c8c4f0810100917t2c903082w5c9ad156043daab4@mail.gmail.com>

Hi,

This doesn't work:
In [19]: spspla.eigen(np.array([[1.0]]),k=1)
---------------------------------------------------------------------------
<type 'exceptions.ValueError'>            Traceback (most recent call last)

/home/james/<ipython console> in <module>()

/usr/lib/python2.5/site-packages/scipy/sparse/linalg/eigen/arpack/arpack.py
in eigen(A, k, M, sigma, which, v0, ncv, maxiter, tol,
return_eigenvectors)
    163         raise ValueError("k must be positive, k=%d"%k)
    164     if k == n:
--> 165         raise ValueError("k must be less than rank(A), k=%d"%k)
    166     if maxiter <= 0:
    167         raise ValueError("maxiter must be positive, maxiter=%d"%maxiter)

<type 'exceptions.ValueError'>: k must be less than rank(A), k=1

It should find the eigenvector [[1.0]] with eigenvalue 1.0.

James


From washakie at gmail.com  Fri Oct 10 12:20:56 2008
From: washakie at gmail.com (John [H2O])
Date: Fri, 10 Oct 2008 09:20:56 -0700 (PDT)
Subject: [SciPy-user] scipy sclicing
In-Reply-To: <ce557a360810100659m23e3c8b9p6a6f5b3011d3bd24@mail.gmail.com>
References: <19917625.post@talk.nabble.com> <19918233.post@talk.nabble.com>
	<ce557a360810100659m23e3c8b9p6a6f5b3011d3bd24@mail.gmail.com>
Message-ID: <19921526.post@talk.nabble.com>




> 
> Fortunately numpy's broadcasting can do it for you,
> repeating each array along a new axis:
> 
> X[ np.array([1,3,5])[:,np.newaxis], np.array([2,4])[np.newaxis,:] ]
> 
> 
> Anne
> 

So perhaps the approach should be:

def slize(X,i,j):
    from numpy import array, newaxis
    X = X[ array([i)[:,newaxis], array(j)[newaxis,:] ]
    return X

Seems a little easier for me to remember!

-- 
View this message in context: http://www.nabble.com/scipy-sclicing-tp19917625p19921526.html
Sent from the Scipy-User mailing list archive at Nabble.com.



From dmitrey.kroshko at scipy.org  Fri Oct 10 13:25:56 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Fri, 10 Oct 2008 20:25:56 +0300
Subject: [SciPy-user] [optimization] 10X speedup example for scipy.optimize
 solver, via using oofun
Message-ID: <48EF9024.4000007@scipy.org>

Hi all,
if anyone is interested here's 10X speedup example for scipy fmin_ncg 
using openopt oofun. Other solvers (from or beyond scipy.optimize) can 
yield similar speedup.

http://openopt.blogspot.com/2008/10/example-of-10x-speedup-for-nlp-via.html

Regards, D.


From wnbell at gmail.com  Fri Oct 10 13:38:50 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Fri, 10 Oct 2008 13:38:50 -0400
Subject: [SciPy-user] Sparse sub-matrix indexing
In-Reply-To: <2b1c8c4f0810100834w2f2023b0jd78241d70bae377c@mail.gmail.com>
References: <2b1c8c4f0810100834w2f2023b0jd78241d70bae377c@mail.gmail.com>
Message-ID: <d05265cb0810101038s6194883ak5afa632c816fa032@mail.gmail.com>

On Fri, Oct 10, 2008 at 11:34 AM, James Philbin <philbinj at gmail.com> wrote:
> Hi,
>
> I've come across this inconsistency with indexing sparse matrices:
> In [3]: A = np.arange(100).reshape(10,10)
> In [4]: B = spsp.csr_matrix(A)
> In [5]: A[np.array([2]),:][:,np.array([2])]
> Out[5]: array([[22]])
> In [6]: B[np.array([2]),:][:,np.array([2])]

Thanks for the report.  It should be fixed in r4792.

btw, you should get a trac account so future reports don't get lost:
http://projects.scipy.org/scipy/scipy

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From contact at pythonxy.com  Fri Oct 10 16:12:33 2008
From: contact at pythonxy.com (Pierre Raybaut)
Date: Fri, 10 Oct 2008 22:12:33 +0200
Subject: [SciPy-user] [ Python(x,y) ] New release : 2.1.2
Message-ID: <48EFB731.10900@pythonxy.com>

Hi all,

As you may already know, Python(x,y) is a free scientific-oriented
Python Distribution based on Qt and Eclipse providing a self-consistent
scientific development environment.

Release 2.1.2 is now available on http://www.pythonxy.com.
(Full Edition, Basic Edition, Light Edition, Custom Edition and Update)

Changes history
Version 2.1.2 (10-09-2008)

     * Added:
           o mercurial 1.0.2: Revision control system
           o MercurialEclipse 1.1.867: Mercurial Eclipse plugin
           o docutils 0.5.0: Text processing system for processing 
plaintext documentation into useful formats, such as HTML or LaTeX 
(includes reStructuredText)
           o jinja 1.2: Sandboxed template engine (provides a 
Django-like non-XML syntax and compiles templates into executable python 
code)
           o pygments 0.11.1: Generic syntax highlighter for general use 
in all kinds of software
     * Updated:
           o Pydev 1.3.22
           o MinGW 3.4.5.4: added GDB
           o xy 1.0.7: Python html help is automatically generated from 
.chm file
     * Corrected:
           o SciPy 0.6.0.1 : tiny update to remove deprecation warnings 
following NumPy latest update
           o IPython 0.9.1.2: post-install install script was not 
executed entirely
           o Issue 25 (PATH environment variable could be corrupted): 
Python(x,y) (main installer), Console2, MinGW, Notepad++, SWIG, GDAL, 
GDCM, OpenCV, PyQt4 and VTK

Regards,
Pierre Raybaut




From matthew.brett at gmail.com  Fri Oct 10 16:13:42 2008
From: matthew.brett at gmail.com (Matthew Brett)
Date: Fri, 10 Oct 2008 13:13:42 -0700
Subject: [SciPy-user] Reference to algorithm for matrix rank
In-Reply-To: <3d375d730810091640i414d0725scd4374f0037b0499@mail.gmail.com>
References: <1e2af89e0810091113o6af11cb7ue23a792e85be7f42@mail.gmail.com>
	<3d375d730810091640i414d0725scd4374f0037b0499@mail.gmail.com>
Message-ID: <1e2af89e0810101313v4692da27k62601d2c3fcaaab2@mail.gmail.com>

Hi,

> You should get the book _Matrix Computations_ by Golub and van Loan.
> You actually want tol to be relative to S.max(), not an absolute
> tolerance. I like this:
>
>  np.sum(S > (S.max() * np.finfo(M.dtype).eps)

Thanks a lot - I'll have a look.

I saw that matlab does something like this:

eps = np.finfo(S.dtype).eps
tol = max(M.shape)*eps*S[0]

but I didn't know why the max(M.shape)...

Best,

Matthew


From matthieu.brucher at gmail.com  Fri Oct 10 16:16:05 2008
From: matthieu.brucher at gmail.com (Matthieu Brucher)
Date: Fri, 10 Oct 2008 22:16:05 +0200
Subject: [SciPy-user] Reference to algorithm for matrix rank
In-Reply-To: <1e2af89e0810101313v4692da27k62601d2c3fcaaab2@mail.gmail.com>
References: <1e2af89e0810091113o6af11cb7ue23a792e85be7f42@mail.gmail.com>
	<3d375d730810091640i414d0725scd4374f0037b0499@mail.gmail.com>
	<1e2af89e0810101313v4692da27k62601d2c3fcaaab2@mail.gmail.com>
Message-ID: <e76aa17f0810101316o1d753084gf8af036583f0746b@mail.gmail.com>

2008/10/10 Matthew Brett <matthew.brett at gmail.com>:
> Hi,
>
>> You should get the book _Matrix Computations_ by Golub and van Loan.
>> You actually want tol to be relative to S.max(), not an absolute
>> tolerance. I like this:
>>
>>  np.sum(S > (S.max() * np.finfo(M.dtype).eps)
>
> Thanks a lot - I'll have a look.
>
> I saw that matlab does something like this:
>
> eps = np.finfo(S.dtype).eps
> tol = max(M.shape)*eps*S[0]
>
> but I didn't know why the max(M.shape)...

The bigger the matrix, the bigger the numerical errors are ?

Matthieu
-- 
Information System Engineer, PhD
Website: http://matthieu-brucher.developpez.com/
Blogs: http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn: http://www.linkedin.com/in/matthieubrucher


From robert.kern at gmail.com  Fri Oct 10 16:45:32 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 10 Oct 2008 15:45:32 -0500
Subject: [SciPy-user] Reference to algorithm for matrix rank
In-Reply-To: <1e2af89e0810101313v4692da27k62601d2c3fcaaab2@mail.gmail.com>
References: <1e2af89e0810091113o6af11cb7ue23a792e85be7f42@mail.gmail.com>
	<3d375d730810091640i414d0725scd4374f0037b0499@mail.gmail.com>
	<1e2af89e0810101313v4692da27k62601d2c3fcaaab2@mail.gmail.com>
Message-ID: <3d375d730810101345q7569a685t1806de466169211e@mail.gmail.com>

On Fri, Oct 10, 2008 at 15:13, Matthew Brett <matthew.brett at gmail.com> wrote:
> Hi,
>
>> You should get the book _Matrix Computations_ by Golub and van Loan.
>> You actually want tol to be relative to S.max(), not an absolute
>> tolerance. I like this:
>>
>>  np.sum(S > (S.max() * np.finfo(M.dtype).eps)
>
> Thanks a lot - I'll have a look.
>
> I saw that matlab does something like this:
>
> eps = np.finfo(S.dtype).eps
> tol = max(M.shape)*eps*S[0]
>
> but I didn't know why the max(M.shape)...

Neither do I. Golub and van Loan define "numerical rank deficiency" as
using tol=eps*S[0] (note that S[0] is the maximum singular value and
thus the 2-norm of the matrix). The thing is, there really isn't one
definition, much like there isn't a single definition of the norm of a
matrix. For example, if your data come from uncertain measurements
with uncertainties greater than floating point epsilon, choosing a
tolerance of about the uncertainty is probably a better idea (the
tolerance may be absolute if the uncertainties are absolute rather
than relative, even). When floating point roundoff is your concern,
then "numerical rank deficiency" is a better concept, but exactly what
the relevant measure of the tolerance is depends on the operations you
intend to do with your matrix. Possibly, the Matlab implementors had
some set of operations in mind. Unfortunately, they don't cite
anything relevant to that point, and my imagination fails to come up
with one.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From alan at ajackson.org  Fri Oct 10 21:14:49 2008
From: alan at ajackson.org (Alan Jackson)
Date: Fri, 10 Oct 2008 20:14:49 -0500
Subject: [SciPy-user] Extrema finding
In-Reply-To: <F47B9FAE-1E8C-41D7-A658-C77DEE0B1536@comcast.net>
References: <bfb379e10810090444i4e435eb8q5f5729005087881b@mail.gmail.com>
	<88e473830810090521qe247f41l9649f3c460e95d38@mail.gmail.com>
	<bfb379e10810090920i290af59dtfdf2c1394ebc4b1f@mail.gmail.com>
	<F47B9FAE-1E8C-41D7-A658-C77DEE0B1536@comcast.net>
Message-ID: <20081010201449.23c1644d@ajackson.org>

On Thu, 9 Oct 2008 13:06:33 -0600
David Bolme <bolme1234 at comcast.net> wrote:

> This is a version for finding extrema in a 2D array.  It requires the  
> ndimage maximum/minimum filters.  For the one dimensional case  
> substitute size=[3] or use maximum_filter1d.  I keep writing this code  
> over and over.  I am surprised that there is not a general purpose  
> extrema finding routine in scipy.
> 
>      def localMax(mat):
>          mx = maximum_filter(mat, size=[3,3])
>          mn = minimum_filter(mat, size=[3,3])
> 	
>          # (mat == mx) true if pixel is equal to the local max
>          # The next computation suppresses responses where
>          # the function is flat.
>          local_maxima = ((mat == mx) & (mat != mn))
> 	
>          # Get the indices of the maxima
>          extrema = nonzero(local_maxima)
>          return extrema
> 
> 
Here's one I wrote a few months back...

def extrema(trace)
    a = np.sign(np.diff(trace))
    zerolocs = np.transpose(np.where(a[1:]+a[0:-1]==0.)).flatten() + 1
    if zerolocs[0] < 1:
        zerolocs = zerolocs[1:]
    if zerolocs[-1]>len(a)-2:
        zerolocs = zerolocs[0:-1]
    return zerolocs

-- 
-----------------------------------------------------------------------
| Alan K. Jackson            | To see a World in a Grain of Sand      |
| alan at ajackson.org          | And a Heaven in a Wild Flower,         |
| www.ajackson.org           | Hold Infinity in the palm of your hand |
| Houston, Texas             | And Eternity in an hour. - Blake       |
-----------------------------------------------------------------------


From nwagner at iam.uni-stuttgart.de  Sat Oct 11 03:35:33 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sat, 11 Oct 2008 09:35:33 +0200
Subject: [SciPy-user] [optimization] 10X speedup example for
 scipy.optimize solver, via using oofun
In-Reply-To: <48EF9024.4000007@scipy.org>
References: <48EF9024.4000007@scipy.org>
Message-ID: <web-112644536@uni-stuttgart.de>

On Fri, 10 Oct 2008 20:25:56 +0300
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> Hi all,
> if anyone is interested here's 10X speedup example for 
>scipy fmin_ncg 
> using openopt oofun. Other solvers (from or beyond 
>scipy.optimize) can 
> yield similar speedup.
> 
> http://openopt.blogspot.com/2008/10/example-of-10x-speedup-for-nlp-via.html
> 
> Regards, D.
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user


python -i 
svn/openopt/scikits/openopt/examples/oofun/speedup.py
Traceback (most recent call last):
   File 
"svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
line 25, in ?
     from scikits.openopt import NLP,  oofun,  oovar
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/__init__.py", 
line 6, in ?
     from oo import LP, NLP, NSP, MILP, QP, NLSP, LSP, 
GLP, LLSP,  MMP, LLAVP
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/oo.py", 
line 6, in ?
     from Kernel.BaseProblem import *
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
line 11, in ?
     from ooIterPrint import ooTextOutput
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/ooIterPrint.py", 
line 6
     'isFeasible': lambda p: ('+' if p.rk<p.contol else 
'-')
                                   ^
SyntaxError: invalid syntax

Nils
  


From dmitrey.kroshko at scipy.org  Sat Oct 11 07:15:52 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Sat, 11 Oct 2008 14:15:52 +0300
Subject: [SciPy-user] [optimization] 10X speedup example for
 scipy.optimize solver, via using oofun
In-Reply-To: <web-112644536@uni-stuttgart.de>
References: <48EF9024.4000007@scipy.org> <web-112644536@uni-stuttgart.de>
Message-ID: <48F08AE8.6000006@scipy.org>

So now Python 2.5 is required (as it is informed in oo install page). Mb 
I could fix it and commit  (mb I'll do this today) but IIRC this is not 
a single place where I use Python 2.5 features.

to run the example, use
iprint = -1(line 23)
and to check the time elapsed add the lines after r=p.solve(...):
print r.elapsed['solver_time']
print r.elapsed['solver_cputime']

Regards, D.

Nils Wagner wrote:
> On Fri, 10 Oct 2008 20:25:56 +0300
>   dmitrey <dmitrey.kroshko at scipy.org> wrote:
>   
>> Hi all,
>> if anyone is interested here's 10X speedup example for 
>> scipy fmin_ncg 
>> using openopt oofun. Other solvers (from or beyond 
>> scipy.optimize) can 
>> yield similar speedup.
>>
>> http://openopt.blogspot.com/2008/10/example-of-10x-speedup-for-nlp-via.html
>>
>> Regards, D.
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>     
>
>
> python -i 
> svn/openopt/scikits/openopt/examples/oofun/speedup.py
> Traceback (most recent call last):
>    File 
> "svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
> line 25, in ?
>      from scikits.openopt import NLP,  oofun,  oovar
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/__init__.py", 
> line 6, in ?
>      from oo import LP, NLP, NSP, MILP, QP, NLSP, LSP, 
> GLP, LLSP,  MMP, LLAVP
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/oo.py", 
> line 6, in ?
>      from Kernel.BaseProblem import *
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
> line 11, in ?
>      from ooIterPrint import ooTextOutput
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/ooIterPrint.py", 
> line 6
>      'isFeasible': lambda p: ('+' if p.rk<p.contol else 
> '-')
>                                    ^
> SyntaxError: invalid syntax
>
> Nils
>   
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From dmitrey.kroshko at scipy.org  Sat Oct 11 07:23:41 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Sat, 11 Oct 2008 14:23:41 +0300
Subject: [SciPy-user] [optimization] 10X speedup example for
 scipy.optimize solver, via using oofun
In-Reply-To: <48F08AE8.6000006@scipy.org>
References: <48EF9024.4000007@scipy.org> <web-112644536@uni-stuttgart.de>
	<48F08AE8.6000006@scipy.org>
Message-ID: <48F08CBD.60405@scipy.org>

I have committed the iterprint changes to svn, you could try now.
Regards, D.

dmitrey wrote:
> Nils Wagner wrote:
>   
>> On Fri, 10 Oct 2008 20:25:56 +0300
>>   dmitrey <dmitrey.kroshko at scipy.org> wrote:
>>   
>>     
>>> Hi all,
>>> if anyone is interested here's 10X speedup example for 
>>> scipy fmin_ncg 
>>> using openopt oofun. Other solvers (from or beyond 
>>> scipy.optimize) can 
>>> yield similar speedup.
>>>
>>> http://openopt.blogspot.com/2008/10/example-of-10x-speedup-for-nlp-via.html
>>>
>>> Regards, D.
>>> _______________________________________________
>>> SciPy-user mailing list
>>> SciPy-user at scipy.org
>>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>>     
>>>       
>> python -i 
>> svn/openopt/scikits/openopt/examples/oofun/speedup.py
>> Traceback (most recent call last):
>>    File 
>> "svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
>> line 25, in ?
>>      from scikits.openopt import NLP,  oofun,  oovar
>>    File 
>> "/usr/lib/python2.4/site-packages/scikits/openopt/__init__.py", 
>> line 6, in ?
>>      from oo import LP, NLP, NSP, MILP, QP, NLSP, LSP, 
>> GLP, LLSP,  MMP, LLAVP
>>    File 
>> "/usr/lib/python2.4/site-packages/scikits/openopt/oo.py", 
>> line 6, in ?
>>      from Kernel.BaseProblem import *
>>    File 
>> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
>> line 11, in ?
>>      from ooIterPrint import ooTextOutput
>>    File 
>> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/ooIterPrint.py", 
>> line 6
>>      'isFeasible': lambda p: ('+' if p.rk<p.contol else 
>> '-')
>>                                    ^
>> SyntaxError: invalid syntax
>>
>> Nils
>>   
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>>
>>
>>   
>>     
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From nwagner at iam.uni-stuttgart.de  Sat Oct 11 11:46:31 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sat, 11 Oct 2008 17:46:31 +0200
Subject: [SciPy-user] openopt ImportError: cannot import name oovar
Message-ID: <web-112656332@uni-stuttgart.de>

python -i 
svn/openopt/scikits/openopt/examples/oofun/speedup.py
Traceback (most recent call last):
   File 
"svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
line 28, in <module>
     from scikits.openopt import NLP,  oofun,  oovar
ImportError: cannot import name oovar

  


From dmitrey.kroshko at scipy.org  Sat Oct 11 15:17:33 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Sat, 11 Oct 2008 22:17:33 +0300
Subject: [SciPy-user] openopt ImportError: cannot import name oovar
In-Reply-To: <web-112656332@uni-stuttgart.de>
References: <web-112656332@uni-stuttgart.de>
Message-ID: <48F0FBCD.3080109@scipy.org>

I checked out openopt from svn and it works for me.
Does anyone else has the problem?

D.

Nils Wagner wrote:
> python -i 
> svn/openopt/scikits/openopt/examples/oofun/speedup.py
> Traceback (most recent call last):
>    File 
> "svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
> line 28, in <module>
>      from scikits.openopt import NLP,  oofun,  oovar
> ImportError: cannot import name oovar
>
>   
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From nwagner at iam.uni-stuttgart.de  Sun Oct 12 04:55:39 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sun, 12 Oct 2008 10:55:39 +0200
Subject: [SciPy-user] openopt ImportError: cannot import name oovar
In-Reply-To: <48F0FBCD.3080109@scipy.org>
References: <web-112656332@uni-stuttgart.de>
 <48F0FBCD.3080109@scipy.org>
Message-ID: <web-112673344@uni-stuttgart.de>

On Sat, 11 Oct 2008 22:17:33 +0300
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> I checked out openopt from svn and it works for me.
> Does anyone else has the problem?
> 
> D.
> 
> Nils Wagner wrote:
>> python -i 
>> svn/openopt/scikits/openopt/examples/oofun/speedup.py
>> Traceback (most recent call last):
>>    File 
>> "svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
>> line 28, in <module>
>>      from scikits.openopt import NLP,  oofun,  oovar
>> ImportError: cannot import name oovar
>>
  
Strange.
I can import oovar on my old laptop (python2.4).

Python 2.4 (#1, Oct 13 2006, 17:13:31)
[GCC 3.3.5 20050117 (prerelease) (SUSE Linux)] on linux2
Type "help", "copyright", "credits" or "license" for more 
information.
>>> from scikits.openopt import oovar
>>>               

Another python2.4 issue is present

  python -i 
svn/openopt/scikits/openopt/examples/oofun/speedup.py
Traceback (most recent call last):
   File 
"svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
line 38, in ?
     r = p.solve(solver)
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
line 201, in solve
     return runProbSolver(self, *args, **kwargs)
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/runProbSolver.py", 
line 102, in runProbSolver
     p.__prepare__()
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
line 385, in __prepare__
     self.__construct_x_from_ooVars__()
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
line 370, in __construct_x_from_ooVars__
     var.__initialize__(self)
   File 
"/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/ooVar.py", 
line 63, in __initialize__
     if any(self.lb > self.ub):
NameError: global name 'any' is not defined

Nils


From dmitrey.kroshko at scipy.org  Sun Oct 12 07:10:43 2008
From: dmitrey.kroshko at scipy.org (dmitrey)
Date: Sun, 12 Oct 2008 14:10:43 +0300
Subject: [SciPy-user] openopt ImportError: cannot import name oovar
In-Reply-To: <web-112673344@uni-stuttgart.de>
References: <web-112656332@uni-stuttgart.de> <48F0FBCD.3080109@scipy.org>
	<web-112673344@uni-stuttgart.de>
Message-ID: <48F1DB33.7030707@scipy.org>

Hi Nils,
try now.

as for the importing issue, maybe you updated svn from

svn/openopt/scikits/openopt/

while it is has to be from svn/openopt/

I'm not fond of the folders structure but it was demanded from all 
scikits (I don't know why less deeply structured folders were not 
organized, it brings lots of inconveniences).

Regards,D.


Nils Wagner wrote:
> On Sat, 11 Oct 2008 22:17:33 +0300
>   dmitrey <dmitrey.kroshko at scipy.org> wrote:
>   
>> I checked out openopt from svn and it works for me.
>> Does anyone else has the problem?
>>
>> D.
>>
>> Nils Wagner wrote:
>>     
>>> python -i 
>>> svn/openopt/scikits/openopt/examples/oofun/speedup.py
>>> Traceback (most recent call last):
>>>    File 
>>> "svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
>>> line 28, in <module>
>>>      from scikits.openopt import NLP,  oofun,  oovar
>>> ImportError: cannot import name oovar
>>>
>>>       
>   
> Strange.
> I can import oovar on my old laptop (python2.4).
>
> Python 2.4 (#1, Oct 13 2006, 17:13:31)
> [GCC 3.3.5 20050117 (prerelease) (SUSE Linux)] on linux2
> Type "help", "copyright", "credits" or "license" for more 
> information.
>   
>>>> from scikits.openopt import oovar
>>>>               
>>>>         
>
> Another python2.4 issue is present
>
>   python -i 
> svn/openopt/scikits/openopt/examples/oofun/speedup.py
> Traceback (most recent call last):
>    File 
> "svn/openopt/scikits/openopt/examples/oofun/speedup.py", 
> line 38, in ?
>      r = p.solve(solver)
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
> line 201, in solve
>      return runProbSolver(self, *args, **kwargs)
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/runProbSolver.py", 
> line 102, in runProbSolver
>      p.__prepare__()
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
> line 385, in __prepare__
>      self.__construct_x_from_ooVars__()
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/BaseProblem.py", 
> line 370, in __construct_x_from_ooVars__
>      var.__initialize__(self)
>    File 
> "/usr/lib/python2.4/site-packages/scikits/openopt/Kernel/ooVar.py", 
> line 63, in __initialize__
>      if any(self.lb > self.ub):
> NameError: global name 'any' is not defined
>
> Nils
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
>   



From nwagner at iam.uni-stuttgart.de  Sun Oct 12 07:30:51 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sun, 12 Oct 2008 13:30:51 +0200
Subject: [SciPy-user] openopt ImportError: cannot import name oovar
In-Reply-To: <48F1DB33.7030707@scipy.org>
References: <web-112656332@uni-stuttgart.de> <48F0FBCD.3080109@scipy.org>
	<web-112673344@uni-stuttgart.de> <48F1DB33.7030707@scipy.org>
Message-ID: <web-112676162@uni-stuttgart.de>

On Sun, 12 Oct 2008 14:10:43 +0300
  dmitrey <dmitrey.kroshko at scipy.org> wrote:
> Hi Nils,
> try now.
> 
> as for the importing issue, maybe you updated svn from
> 
> svn/openopt/scikits/openopt/
> 
> while it is has to be from svn/openopt/
> 
> I'm not fond of the folders structure but it was 
>demanded from all 
> scikits (I don't know why less deeply structured folders 
>were not 
> organized, it brings lots of inconveniences).
> 
> Regards,D.


Hi Dmitrey,

I use
svn co http://svn.scipy.org/svn/scikits/trunk/openopt 
openopt

in $HOME/svn

to checkout openopt.

Nils

  


From nwagner at iam.uni-stuttgart.de  Sun Oct 12 07:37:05 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Sun, 12 Oct 2008 13:37:05 +0200
Subject: [SciPy-user] openopt ImportError: cannot import name oovar
In-Reply-To: <48F1DB33.7030707@scipy.org>
References: <web-112656332@uni-stuttgart.de> <48F0FBCD.3080109@scipy.org>
	<web-112673344@uni-stuttgart.de> <48F1DB33.7030707@scipy.org>
Message-ID: <web-112676381@uni-stuttgart.de>

Hi Dmitrey,

The example speedup.py works with r1537. Thank you.

python -i 
svn/openopt/scikits/openopt/examples/oofun/speedup.py
-----------------------------------------------------
solver: scipy_ncg   problem: unnamed   goal: minimum
  iter    objFunVal
     0  2.228e+06
     5  1.142e+06
    10  6.139e+02
    15  4.909e+02
    20  4.888e+02
    25  4.869e+02
    28  4.862e+02
istop:  1000
Solver:   Time Elapsed = 5.95   CPU Time Elapsed = 5.91
objFunValue: 486.19658
evals f:  38562 evals of costly func g: 567
-----------------------------------------------------
solver: scipy_ncg   problem: unnamed   goal: minimum
  iter    objFunVal
     0  2.228e+06
     5  4.982e+02

    10  4.938e+02
    15  4.912e+02
    20  4.891e+02
    25  4.871e+02
    29  4.861e+02
istop:  1000
Solver:   Time Elapsed = 42.5   CPU Time Elapsed = 40.3
objFunValue: 486.11448
evals f:  24235 evals of costly func g: 24235

Cheers,
           Nils


From kwmsmith at gmail.com  Sun Oct 12 14:45:57 2008
From: kwmsmith at gmail.com (Kurt Smith)
Date: Sun, 12 Oct 2008 18:45:57 +0000
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
Message-ID: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>

Hi list,

I've recently acquired a new Ubuntu system, and unfortunately some 3rd
party software I'd like to use won't work under the 64 bit version of
the OS, yet.  It's not a deal breaker, however, and I'm willing to
work around it as long as there are compelling reasons to use 64 bit
for numerical work, specifically under scipy, etc.  I'm not well
versed in the low-level pros and cons of 64 bit vs 32 bit, esp. as to
how it affects scipy performance.  Can you enlighten me?  I"m on an
Intel Core 2 Duo, 3.16 GHz, E8500, 4 GB RAM.  Will using 64 bit yield
great benefits running scipy/numpy code automatically, or will it
require fine-tuning on my part?  Will 64 bit yield benefits besides
having a larger address space (besides the fact that my 4 GB RAM is
still addressable by 32 bits, anyway)?


Thanks for your input,

Kurt


From wnbell at gmail.com  Sun Oct 12 15:14:18 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Sun, 12 Oct 2008 15:14:18 -0400
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
Message-ID: <d05265cb0810121214o1672b32bv9c69e76a4dd53c89@mail.gmail.com>

On Sun, Oct 12, 2008 at 2:45 PM, Kurt Smith <kwmsmith at gmail.com> wrote:
>
> for numerical work, specifically under scipy, etc.  I'm not well
> versed in the low-level pros and cons of 64 bit vs 32 bit, esp. as to
> how it affects scipy performance.  Can you enlighten me?  I"m on an
> Intel Core 2 Duo, 3.16 GHz, E8500, 4 GB RAM.  Will using 64 bit yield
> great benefits running scipy/numpy code automatically, or will it
> require fine-tuning on my part?  Will 64 bit yield benefits besides
> having a larger address space (besides the fact that my 4 GB RAM is
> still addressable by 32 bits, anyway)?
>

I wouldn't expect any significant performance difference.  The x86-64
architecture does bring a few improvements such as more registers and
(guaranteed?) SSE2 support, but the benefit of these depends on the
nature of the application.
http://en.wikipedia.org/wiki/X86-64#Architectural_features

Something in the neighborhood of 20% faster for general purpose codes
should be typical.  OTOH I've found that some pure Python code that
makes heavy use of dictionaries runs slower, possibly owing to the
fact that 64-bit pointers reduce the effective CPU cache size.

Also, you wouldn't be able to address *all* 4GB RAM in a 32-bit OS.
Both Windows and Linux reserve some part of the upper addresses for
the OS, so a process will be limited to something less (3GB to 3.5GB
IIRC).

Keep in mind that 64-bit Linux is approximately 2x cooler than 32-bit
Linux, so the choice is easy IMO.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From warren.weckesser at gmail.com  Sun Oct 12 16:02:48 2008
From: warren.weckesser at gmail.com (Warren Weckesser)
Date: Sun, 12 Oct 2008 15:02:48 -0500
Subject: [SciPy-user] f2py: problem with PUBLIC access spec of a derived type
Message-ID: <114880320810121302w5b970ba7n4f8bab188af1a1d3@mail.gmail.com>

Hi,

I am learning about f2py, with the goal of wrapping a delay differential
equation solver that is written in Fortran 90.  The solver is Shampine and
Thompson's DDE_SOLVER.  Here is what I get when I run f2py:

------------
$ f2py dde_solver_m_unix.f90 -h tmp.pyf
Reading fortran codes...
        Reading file 'dde_solver_m_unix.f90' (format:free)
Line #316 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_SOL"
        analyzeline: No name/args pattern found for line.
Line #331 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_OPTS"
        analyzeline: No name/args pattern found for line.
Line #340 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_INT"
        analyzeline: No name/args pattern found for line.
rmbadname1: Replacing "index" with "index_bn".
rmbadname1: Replacing "index" with "index_bn".
rmbadname1: Replacing "index" with "index_bn".
<snip>
------------

Why is f2py complaining about the PUBLIC access spec for the derived types?

Regards,

Warren
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081012/4a094aed/attachment.html>

From kwmsmith at gmail.com  Sun Oct 12 17:04:34 2008
From: kwmsmith at gmail.com (Kurt Smith)
Date: Sun, 12 Oct 2008 21:04:34 +0000
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <d05265cb0810121214o1672b32bv9c69e76a4dd53c89@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<d05265cb0810121214o1672b32bv9c69e76a4dd53c89@mail.gmail.com>
Message-ID: <c266a7930810121404o5fb82058s17d65ee21683a818@mail.gmail.com>

On Sun, Oct 12, 2008 at 7:14 PM, Nathan Bell <wnbell at gmail.com> wrote:
> On Sun, Oct 12, 2008 at 2:45 PM, Kurt Smith <kwmsmith at gmail.com> wrote:
>>
>> for numerical work, specifically under scipy, etc.  I'm not well
>> versed in the low-level pros and cons of 64 bit vs 32 bit, esp. as to
>> how it affects scipy performance.  Can you enlighten me?  I"m on an
>> Intel Core 2 Duo, 3.16 GHz, E8500, 4 GB RAM.  Will using 64 bit yield
>> great benefits running scipy/numpy code automatically, or will it
>> require fine-tuning on my part?  Will 64 bit yield benefits besides
>> having a larger address space (besides the fact that my 4 GB RAM is
>> still addressable by 32 bits, anyway)?
>>
>
> I wouldn't expect any significant performance difference.  The x86-64
> architecture does bring a few improvements such as more registers and
> (guaranteed?) SSE2 support, but the benefit of these depends on the
> nature of the application.
> http://en.wikipedia.org/wiki/X86-64#Architectural_features
>
> Something in the neighborhood of 20% faster for general purpose codes
> should be typical.  OTOH I've found that some pure Python code that
> makes heavy use of dictionaries runs slower, possibly owing to the
> fact that 64-bit pointers reduce the effective CPU cache size.

Is this 20% from your own experience, or are there timings out there?
I've looked around, but the comparisons are pretty old and made when
very few 64 bit apps were around, so 32 had the day.  Anyone aware of
comparisons between 64 and 32 bit performance for numerical codes?


> Also, you wouldn't be able to address *all* 4GB RAM in a 32-bit OS.
> Both Windows and Linux reserve some part of the upper addresses for
> the OS, so a process will be limited to something less (3GB to 3.5GB
> IIRC).
>
> Keep in mind that 64-bit Linux is approximately 2x cooler than 32-bit
> Linux, so the choice is easy IMO.

This temperature difference is really surprising to me, and I'm not
able to find anything that talks about it in detail.  Anything more
you can say as to why this would be the case?

Thanks,

Kurt


From warren.weckesser at gmail.com  Sun Oct 12 17:16:20 2008
From: warren.weckesser at gmail.com (Warren Weckesser)
Date: Sun, 12 Oct 2008 16:16:20 -0500
Subject: [SciPy-user] f2py: problem with PUBLIC access spec of a derived
	type
In-Reply-To: <114880320810121302w5b970ba7n4f8bab188af1a1d3@mail.gmail.com>
References: <114880320810121302w5b970ba7n4f8bab188af1a1d3@mail.gmail.com>
Message-ID: <114880320810121416y567efd4eg7aebc93cec9cc9f3@mail.gmail.com>

Hi again,

On Sun, Oct 12, 2008 at 3:02 PM, Warren Weckesser <
warren.weckesser at gmail.com> wrote:

> Hi,
>
> I am learning about f2py, with the goal of wrapping a delay differential
> equation solver that is written in Fortran 90.  The solver is Shampine and
> Thompson's DDE_SOLVER.  Here is what I get when I run f2py:
>
> ------------
> $ f2py dde_solver_m_unix.f90 -h tmp.pyf
> Reading fortran codes...
>         Reading file 'dde_solver_m_unix.f90' (format:free)
> Line #316 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_SOL"
>         analyzeline: No name/args pattern found for line.
> Line #331 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_OPTS"
>         analyzeline: No name/args pattern found for line.
> Line #340 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_INT"
>         analyzeline: No name/args pattern found for line.
> rmbadname1: Replacing "index" with "index_bn".
> rmbadname1: Replacing "index" with "index_bn".
> rmbadname1: Replacing "index" with "index_bn".
> <snip>
> ------------
>
> Why is f2py complaining about the PUBLIC access spec for the derived types?
>

I should have added that the problem appears to be the PUBLIC spec because
if I remove the PUBLIC spec, so, for example, line 316 is  "  TYPE DDE_SOL",
then I don't get the "analyzeline" errors.

Warren
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081012/6ab65e8e/attachment.html>

From warren.weckesser at gmail.com  Sun Oct 12 19:39:39 2008
From: warren.weckesser at gmail.com (Warren Weckesser)
Date: Sun, 12 Oct 2008 18:39:39 -0500
Subject: [SciPy-user] f2py: problem with PUBLIC access spec of a derived
	type
In-Reply-To: <114880320810121416y567efd4eg7aebc93cec9cc9f3@mail.gmail.com>
References: <114880320810121302w5b970ba7n4f8bab188af1a1d3@mail.gmail.com>
	<114880320810121416y567efd4eg7aebc93cec9cc9f3@mail.gmail.com>
Message-ID: <114880320810121639j39220781k536f860e2787418@mail.gmail.com>

Answering my own question:  according to the f2py FAQ at
http://cens.ioc.ee/projects/f2py2e/FAQ.html, derived types are not supported
in F90 code.

Warren


On Sun, Oct 12, 2008 at 4:16 PM, Warren Weckesser <
warren.weckesser at gmail.com> wrote:

> Hi again,
>
> On Sun, Oct 12, 2008 at 3:02 PM, Warren Weckesser <
> warren.weckesser at gmail.com> wrote:
>
>> Hi,
>>
>> I am learning about f2py, with the goal of wrapping a delay differential
>> equation solver that is written in Fortran 90.  The solver is Shampine and
>> Thompson's DDE_SOLVER.  Here is what I get when I run f2py:
>>
>> ------------
>> $ f2py dde_solver_m_unix.f90 -h tmp.pyf
>> Reading fortran codes...
>>         Reading file 'dde_solver_m_unix.f90' (format:free)
>> Line #316 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_SOL"
>>         analyzeline: No name/args pattern found for line.
>> Line #331 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_OPTS"
>>         analyzeline: No name/args pattern found for line.
>> Line #340 in dde_solver_m_unix.f90:"  TYPE, PUBLIC :: DDE_INT"
>>         analyzeline: No name/args pattern found for line.
>> rmbadname1: Replacing "index" with "index_bn".
>> rmbadname1: Replacing "index" with "index_bn".
>> rmbadname1: Replacing "index" with "index_bn".
>> <snip>
>> ------------
>>
>> Why is f2py complaining about the PUBLIC access spec for the derived
>> types?
>>
>
> I should have added that the problem appears to be the PUBLIC spec because
> if I remove the PUBLIC spec, so, for example, line 316 is  "  TYPE DDE_SOL",
> then I don't get the "analyzeline" errors.
>
> Warren
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081012/accd8863/attachment.html>

From robert.kern at gmail.com  Mon Oct 13 01:53:19 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 13 Oct 2008 00:53:19 -0500
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <c266a7930810121404o5fb82058s17d65ee21683a818@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<d05265cb0810121214o1672b32bv9c69e76a4dd53c89@mail.gmail.com>
	<c266a7930810121404o5fb82058s17d65ee21683a818@mail.gmail.com>
Message-ID: <3d375d730810122253v7480cbbdwd89fae8c01e394d4@mail.gmail.com>

On Sun, Oct 12, 2008 at 16:04, Kurt Smith <kwmsmith at gmail.com> wrote:
> On Sun, Oct 12, 2008 at 7:14 PM, Nathan Bell <wnbell at gmail.com> wrote:

>> Keep in mind that 64-bit Linux is approximately 2x cooler than 32-bit
>> Linux, so the choice is easy IMO.
>
> This temperature difference is really surprising to me, and I'm not
> able to find anything that talks about it in detail.  Anything more
> you can say as to why this would be the case?

He didn't mean "cooler" as in, "Yesterday was cooler than today," but
rather "cooler" as in, "James Dean was cooler than I will ever be."

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From philbinj at gmail.com  Mon Oct 13 10:04:32 2008
From: philbinj at gmail.com (James Philbin)
Date: Mon, 13 Oct 2008 15:04:32 +0100
Subject: [SciPy-user] scipy.sparse: coo_matrix ignores sum_duplicates=False
Message-ID: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>

I've filed this as trac #754, repeated here for visibility.

---
Running scipy version 0.7.0.dev4763

coo_matrix.tocsr + tocsc both ignore the sum_duplicates parameter:

In [1]: from numpy import *
In [2]: from scipy.sparse import *
In [3]: data = array([1,1,1,1,1,1,1])
In [4]: row  = array([0,0,1,3,1,0,0])
In [5]: col  = array([0,2,1,3,1,0,0])
In [6]: A = coo_matrix( (data,(row,col)), shape=(4,4))
In [8]: A.tocsr(sum_duplicates=False).todense()
Out[8]:
matrix([[3, 0, 1, 0],
        [0, 2, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 1]])
In [9]: A.tocsc(sum_duplicates=False).todense()
Out[9]:
matrix([[3, 0, 1, 0],
        [0, 2, 0, 0],
        [0, 0, 0, 0],
        [0, 0, 0, 1]])
---

Thanks,
James


From kwmsmith at gmail.com  Mon Oct 13 10:33:59 2008
From: kwmsmith at gmail.com (Kurt Smith)
Date: Mon, 13 Oct 2008 09:33:59 -0500
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <3d375d730810122253v7480cbbdwd89fae8c01e394d4@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<d05265cb0810121214o1672b32bv9c69e76a4dd53c89@mail.gmail.com>
	<c266a7930810121404o5fb82058s17d65ee21683a818@mail.gmail.com>
	<3d375d730810122253v7480cbbdwd89fae8c01e394d4@mail.gmail.com>
Message-ID: <c266a7930810130733o4c68aeb7p2d7a565e9e2d786f@mail.gmail.com>

On Mon, Oct 13, 2008 at 12:53 AM, Robert Kern <robert.kern at gmail.com> wrote:
> On Sun, Oct 12, 2008 at 16:04, Kurt Smith <kwmsmith at gmail.com> wrote:
>> On Sun, Oct 12, 2008 at 7:14 PM, Nathan Bell <wnbell at gmail.com> wrote:
>
>>> Keep in mind that 64-bit Linux is approximately 2x cooler than 32-bit
>>> Linux, so the choice is easy IMO.
>>
>> This temperature difference is really surprising to me, and I'm not
>> able to find anything that talks about it in detail.  Anything more
>> you can say as to why this would be the case?
>
> He didn't mean "cooler" as in, "Yesterday was cooler than today," but
> rather "cooler" as in, "James Dean was cooler than I will ever be.

Thanks -- you saved me from a rather long wild goose chase.


From wnbell at gmail.com  Mon Oct 13 10:35:23 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Mon, 13 Oct 2008 10:35:23 -0400
Subject: [SciPy-user] scipy.sparse: coo_matrix ignores
	sum_duplicates=False
In-Reply-To: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>
References: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>
Message-ID: <d05265cb0810130735p18d8c237ybe8fa54163a35317@mail.gmail.com>

On Mon, Oct 13, 2008 at 10:04 AM, James Philbin <philbinj at gmail.com> wrote:
> I've filed this as trac #754, repeated here for visibility.
>
> ---
> Running scipy version 0.7.0.dev4763
>
> coo_matrix.tocsr + tocsc both ignore the sum_duplicates parameter:
>
> In [1]: from numpy import *
> In [2]: from scipy.sparse import *
> In [3]: data = array([1,1,1,1,1,1,1])
> In [4]: row  = array([0,0,1,3,1,0,0])
> In [5]: col  = array([0,2,1,3,1,0,0])
> In [6]: A = coo_matrix( (data,(row,col)), shape=(4,4))
> In [8]: A.tocsr(sum_duplicates=False).todense()
> Out[8]:
> matrix([[3, 0, 1, 0],
>        [0, 2, 0, 0],
>        [0, 0, 0, 0],
>        [0, 0, 0, 1]])
> In [9]: A.tocsc(sum_duplicates=False).todense()
> Out[9]:
> matrix([[3, 0, 1, 0],
>        [0, 2, 0, 0],
>        [0, 0, 0, 0],
>        [0, 0, 0, 1]])

Hi James,

Note that CSR.todense() implicitly sums duplicate entries (it's
essentially zeros((N,M)) += A).  You should find that the CSR
representation *does* contain the duplicate entries:

In [1]: from numpy import *
In [2]: from scipy.sparse import *
In [3]: data = array([1,1,1,1,1,1,1])
In [4]: row  = array([0,0,1,3,1,0,0])
In [5]: col  = array([0,2,1,3,1,0,0])
In [6]: A = coo_matrix( (data,(row,col)), shape=(4,4))
In [7]: B = A.tocsr(sum_duplicates=False)
In [8]: B.indptr
Out[8]: array([0, 4, 6, 6, 7], dtype=int32)
In [9]: B.indices
Out[9]: array([0, 2, 0, 0, 1, 1, 3], dtype=int32)
In [10]: B.data
Out[10]: array([1, 1, 1, 1, 1, 1, 1])
In [11]: B
Out[11]:
<4x4 sparse matrix of type '<type 'numpy.int64'>'
	with 7 stored elements in Compressed Sparse Row format>

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From wnbell at gmail.com  Mon Oct 13 10:41:31 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Mon, 13 Oct 2008 10:41:31 -0400
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <c266a7930810130733o4c68aeb7p2d7a565e9e2d786f@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<d05265cb0810121214o1672b32bv9c69e76a4dd53c89@mail.gmail.com>
	<c266a7930810121404o5fb82058s17d65ee21683a818@mail.gmail.com>
	<3d375d730810122253v7480cbbdwd89fae8c01e394d4@mail.gmail.com>
	<c266a7930810130733o4c68aeb7p2d7a565e9e2d786f@mail.gmail.com>
Message-ID: <d05265cb0810130741r6da64215n3169bbaceac0640c@mail.gmail.com>

On Mon, Oct 13, 2008 at 10:33 AM, Kurt Smith <kwmsmith at gmail.com> wrote:
>>
>> He didn't mean "cooler" as in, "Yesterday was cooler than today," but
>> rather "cooler" as in, "James Dean was cooler than I will ever be.
>
> Thanks -- you saved me from a rather long wild goose chase.

Yeah, sorry about that. I should have terminated that statement with
an emoticon :)

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From wnbell at gmail.com  Mon Oct 13 10:43:53 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Mon, 13 Oct 2008 10:43:53 -0400
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <c266a7930810121404o5fb82058s17d65ee21683a818@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<d05265cb0810121214o1672b32bv9c69e76a4dd53c89@mail.gmail.com>
	<c266a7930810121404o5fb82058s17d65ee21683a818@mail.gmail.com>
Message-ID: <d05265cb0810130743g95a3a0bg618dbe01a904ab67@mail.gmail.com>

On Sun, Oct 12, 2008 at 5:04 PM, Kurt Smith <kwmsmith at gmail.com> wrote:
>
> Is this 20% from your own experience, or are there timings out there?
> I've looked around, but the comparisons are pretty old and made when
> very few 64 bit apps were around, so 32 had the day.  Anyone aware of
> comparisons between 64 and 32 bit performance for numerical codes?
>

I've never done a proper study of the matter, so that's just my take
from benchmarks and anecdotes I've read.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From philbinj at gmail.com  Mon Oct 13 10:52:51 2008
From: philbinj at gmail.com (James Philbin)
Date: Mon, 13 Oct 2008 15:52:51 +0100
Subject: [SciPy-user] scipy.sparse: coo_matrix ignores
	sum_duplicates=False
In-Reply-To: <d05265cb0810130735p18d8c237ybe8fa54163a35317@mail.gmail.com>
References: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>
	<d05265cb0810130735p18d8c237ybe8fa54163a35317@mail.gmail.com>
Message-ID: <2b1c8c4f0810130752w138837fan9ef20c59204a1eb2@mail.gmail.com>

Hi,

> Note that CSR.todense() implicitly sums duplicate entries (it's
> essentially zeros((N,M)) += A).  You should find that the CSR
> representation *does* contain the duplicate entries:
Hmm, I see. This is quite subtle (+ suprising) as to all intents and
purposes the csr_matrix behaves as if the duplicates had been summed
whether or not sum_duplicates=True or False. The parameter name
probably needs to be changed and/or something said in the docstring.
What I was actually looking for was a way for duplicates to be
ignored, which i've found with dok_matrix.

Thanks a lot,
James


From peridot.faceted at gmail.com  Mon Oct 13 11:25:15 2008
From: peridot.faceted at gmail.com (Anne Archibald)
Date: Mon, 13 Oct 2008 11:25:15 -0400
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
Message-ID: <ce557a360810130825q1fbad70i7fbed40f868daa16@mail.gmail.com>

2008/10/12 Kurt Smith <kwmsmith at gmail.com>:

> I've recently acquired a new Ubuntu system, and unfortunately some 3rd
> party software I'd like to use won't work under the 64 bit version of
> the OS, yet.  It's not a deal breaker, however, and I'm willing to
> work around it as long as there are compelling reasons to use 64 bit
> for numerical work, specifically under scipy, etc.  I'm not well
> versed in the low-level pros and cons of 64 bit vs 32 bit, esp. as to
> how it affects scipy performance.  Can you enlighten me?  I"m on an
> Intel Core 2 Duo, 3.16 GHz, E8500, 4 GB RAM.  Will using 64 bit yield
> great benefits running scipy/numpy code automatically, or will it
> require fine-tuning on my part?  Will 64 bit yield benefits besides
> having a larger address space (besides the fact that my 4 GB RAM is
> still addressable by 32 bits, anyway)?

I think memory is the only major reason to prefer 64 bits. The speed
difference will probably be minimal. But having a 64-bit address space
means that if your applications get too big for RAM they just slow
down, as opposed to crashing and forcing you to rewrite them. You will
also find, I think, that running in 32-bit mode you cannot access the
last gigabyte or so of physical RAM. This can make a big difference
too.

Anne


From mhearne at usgs.gov  Mon Oct 13 13:12:00 2008
From: mhearne at usgs.gov (Michael Hearne)
Date: Mon, 13 Oct 2008 11:12:00 -0600
Subject: [SciPy-user] Results of squeeze function on a [1,1] size array
In-Reply-To: <48E14401.7030600@ru.nl>
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
	<48E14401.7030600@ru.nl>
Message-ID: <48F38160.4090705@usgs.gov>

Using numpy version 1.1.0.dev5077
Using scipy version 0.7.0.dev4174

on Mac OS X

If I do the following:
x = zeros([1,1])
x[0] = 3.4756
y = squeeze(x)

What is y?  When I print or do arithmetic with y, it seems like it's a 
scalar.   However, the type() function seems to indicate that it's a  
numpy.ndarray,  but with 0 dimensionality.

Questions:
How do I detect when I'm in this state?
How can I convert this from a (sort of) scalar into an array with length 
of one?

--Mike


-- 
------------------------------------------------------
Michael Hearne
mhearne at usgs.gov
(303) 273-8620
USGS National Earthquake Information Center
1711 Illinois St. Golden CO 80401
Senior Software Engineer
Synergetics, Inc.
------------------------------------------------------



From robert.kern at gmail.com  Mon Oct 13 13:16:10 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 13 Oct 2008 12:16:10 -0500
Subject: [SciPy-user] Results of squeeze function on a [1,1] size array
In-Reply-To: <48F38160.4090705@usgs.gov>
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>
	<48E14401.7030600@ru.nl> <48F38160.4090705@usgs.gov>
Message-ID: <3d375d730810131016h1296a069id1672ea1e2bfce72@mail.gmail.com>

On Mon, Oct 13, 2008 at 12:12, Michael Hearne <mhearne at usgs.gov> wrote:
> Using numpy version 1.1.0.dev5077
> Using scipy version 0.7.0.dev4174
>
> on Mac OS X
>
> If I do the following:
> x = zeros([1,1])
> x[0] = 3.4756
> y = squeeze(x)
>
> What is y?  When I print or do arithmetic with y, it seems like it's a
> scalar.

Why do you say that? It looks like a 0-dim array to me.

In [1]: from numpy import *

In [2]: x = zeros([1,1])

In [3]: x
Out[3]: array([[ 0.]])

In [4]: x[0] = 3.4756

In [5]: x
Out[5]: array([[ 3.4756]])

In [6]: y = squeeze(x)

In [7]: y
Out[7]: array(3.4756)

In [8]: y.shape
Out[8]: ()

> However, the type() function seems to indicate that it's a
> numpy.ndarray,  but with 0 dimensionality.

Yup.

> Questions:
> How do I detect when I'm in this state?

isinstance(y, numpy.ndarray) and y.shape == ()

> How can I convert this from a (sort of) scalar into an array with length
> of one?

atleast_1d(y)

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From cdcasey at gmail.com  Mon Oct 13 14:09:27 2008
From: cdcasey at gmail.com (chris)
Date: Mon, 13 Oct 2008 13:09:27 -0500
Subject: [SciPy-user] scipy 0.6.0 build failing
Message-ID: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>

I'm trying to build scipy 0.6.0 on RHEL 3, and am getting the following failure:


g77:f77: scipy/fftpack/dfftpack/zfftf1.f
/tmp/cceGs6VT.s: Assembler messages:
/tmp/cceGs6VT.s:598: Error: suffix or operands invalid for `movd'
/tmp/cceGs6VT.s:2994: Error: suffix or operands invalid for `movd'
/tmp/cceGs6VT.s: Assembler messages:
/tmp/cceGs6VT.s:598: Error: suffix or operands invalid for `movd'
/tmp/cceGs6VT.s:2994: Error: suffix or operands invalid for `movd'
error: Command "/usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2
-funroll-loops -march=i686 -mmmx -msse2 -msse -fomit-frame-pointer
-malign-double -c -c scipy/fftpack/dfftpack/zfftf1.f -o
build/temp.linux-i686-2.5/scipy/fftpack/dfftpack/zfftf1.o" failed with
exit status 1


g77 version 3.2.3
gcc version 3.2.3
as version 2.14.90.0.4


From mhearne at usgs.gov  Mon Oct 13 14:31:11 2008
From: mhearne at usgs.gov (Michael Hearne)
Date: Mon, 13 Oct 2008 12:31:11 -0600
Subject: [SciPy-user] Results of squeeze function on a [1,1] size array
In-Reply-To: <3d375d730810131016h1296a069id1672ea1e2bfce72@mail.gmail.com>
References: <0191F372-8348-42DC-AB33-61A2833157A3@gmail.com>	
	<48E14401.7030600@ru.nl> <48F38160.4090705@usgs.gov>
	<3d375d730810131016h1296a069id1672ea1e2bfce72@mail.gmail.com>
Message-ID: <48F393EF.5020708@usgs.gov>

Didn't know there was such a thing as a 0-dim array.  It seems to 
_behave_ like a scalar!

Problem solved...

Thanks,

Mike

Robert Kern wrote:
> On Mon, Oct 13, 2008 at 12:12, Michael Hearne <mhearne at usgs.gov> wrote:
>   
>> Using numpy version 1.1.0.dev5077
>> Using scipy version 0.7.0.dev4174
>>
>> on Mac OS X
>>
>> If I do the following:
>> x = zeros([1,1])
>> x[0] = 3.4756
>> y = squeeze(x)
>>
>> What is y?  When I print or do arithmetic with y, it seems like it's a
>> scalar.
>>     
>
> Why do you say that? It looks like a 0-dim array to me.
>
> In [1]: from numpy import *
>
> In [2]: x = zeros([1,1])
>
> In [3]: x
> Out[3]: array([[ 0.]])
>
> In [4]: x[0] = 3.4756
>
> In [5]: x
> Out[5]: array([[ 3.4756]])
>
> In [6]: y = squeeze(x)
>
> In [7]: y
> Out[7]: array(3.4756)
>
> In [8]: y.shape
> Out[8]: ()
>
>   
>> However, the type() function seems to indicate that it's a
>> numpy.ndarray,  but with 0 dimensionality.
>>     
>
> Yup.
>
>   
>> Questions:
>> How do I detect when I'm in this state?
>>     
>
> isinstance(y, numpy.ndarray) and y.shape == ()
>
>   
>> How can I convert this from a (sort of) scalar into an array with length
>> of one?
>>     
>
> atleast_1d(y)
>
>   

-- 
------------------------------------------------------
Michael Hearne
mhearne at usgs.gov
(303) 273-8620
USGS National Earthquake Information Center
1711 Illinois St. Golden CO 80401
Senior Software Engineer
Synergetics, Inc.
------------------------------------------------------



From wnbell at gmail.com  Mon Oct 13 15:30:29 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Mon, 13 Oct 2008 15:30:29 -0400
Subject: [SciPy-user] scipy.sparse: coo_matrix ignores
	sum_duplicates=False
In-Reply-To: <2b1c8c4f0810130752w138837fan9ef20c59204a1eb2@mail.gmail.com>
References: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>
	<d05265cb0810130735p18d8c237ybe8fa54163a35317@mail.gmail.com>
	<2b1c8c4f0810130752w138837fan9ef20c59204a1eb2@mail.gmail.com>
Message-ID: <d05265cb0810131230l367b30c2ha4a9b0815144a64@mail.gmail.com>

On Mon, Oct 13, 2008 at 10:52 AM, James Philbin <philbinj at gmail.com> wrote:
>
> Hmm, I see. This is quite subtle (+ suprising) as to all intents and
> purposes the csr_matrix behaves as if the duplicates had been summed
> whether or not sum_duplicates=True or False. The parameter name
> probably needs to be changed and/or something said in the docstring.
> What I was actually looking for was a way for duplicates to be
> ignored, which i've found with dok_matrix.
>

By "ignored" do you mean that you want only the first or last value to be used?

Summing duplicates when converting COO->CSR is fairly common (e.g.
UMFPACK does it) and quite useful if you're assembling FEM matrices.
Furthermore, regarding duplicate entries as parts of a sum is
necessary if one wants to maintain consistency with matrix-vector
multiplication (i.e A*x == A.tocsr() * x).  In theory you could change
this as well, but it would be *very* costly.

FYI, others have expressed an interest more general accumulation methods:
http://thread.gmane.org/gmane.comp.python.scientific.devel/7667

I'll think about how to implement this.  It should be straightfoward
to do in pure numpy, but I'd want it to be fast for the common cases
Viral listed in the message above.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From philbinj at gmail.com  Mon Oct 13 18:33:45 2008
From: philbinj at gmail.com (James Philbin)
Date: Mon, 13 Oct 2008 23:33:45 +0100
Subject: [SciPy-user] scipy.sparse: coo_matrix ignores
	sum_duplicates=False
In-Reply-To: <d05265cb0810131230l367b30c2ha4a9b0815144a64@mail.gmail.com>
References: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>
	<d05265cb0810130735p18d8c237ybe8fa54163a35317@mail.gmail.com>
	<2b1c8c4f0810130752w138837fan9ef20c59204a1eb2@mail.gmail.com>
	<d05265cb0810131230l367b30c2ha4a9b0815144a64@mail.gmail.com>
Message-ID: <2b1c8c4f0810131533x430b6254x140691ce2a934942@mail.gmail.com>

> By "ignored" do you mean that you want only the first or last value to be used?
My use case is perhaps a bit non-standard. I'm approximately computing
a large pairwise similarity matrix distributed across multiple
processes. The algorithm will sometimes output the same pairwise
distance more than once, so all the subsequent values will be the
same. I think dok_matrix is fine for my needs. BTW, i've found that
__setitem__ is v slow for dok_matrix. Is this just because of the
checks which are made? Using dict.__setitem__(mat, (r,c), val) is
about an order of magnitude faster.

> Summing duplicates when converting COO->CSR is fairly common (e.g.
> UMFPACK does it) and quite useful if you're assembling FEM matrices.
> Furthermore, regarding duplicate entries as parts of a sum is
> necessary if one wants to maintain consistency with matrix-vector
> multiplication (i.e A*x == A.tocsr() * x).  In theory you could change
> this as well, but it would be *very* costly.
I'm not arguing that summing duplicate entries is not desirable. I'm
just arguing that a function which reads .tocsr(sum_duplicates=False)
and then sums the duplicates implicitly is misnamed.

> FYI, others have expressed an interest more general accumulation methods:
> http://thread.gmane.org/gmane.comp.python.scientific.devel/7667
This is never something i've needed, but I agree it could be useful.

Thanks,
James


From mgb11 at cornell.edu  Mon Oct 13 20:22:26 2008
From: mgb11 at cornell.edu (Marc Berthoud)
Date: Mon, 13 Oct 2008 19:22:26 -0500
Subject: [SciPy-user] Record Array: How to add a column?
Message-ID: <48F3E642.4040001@cornell.edu>

What is the best way to add a column to a record array?

Example:
Add to this array
  recarray( [ ( 'alpha' , 1 ) , ( 'beta ' , 2 ) , ( 'gamma' , 3 ) ] ,
    dtype = [ ( 'NAMES' , '|S5' ) , ( 'NR' , '>i2' ) ] )
the following column
   array( [ 3.1 ,  3.5 ,  8.1 ] )
to make that array
  recarray( [ ( 'alpha' , 1, 3.1 ), ( 'beta ' , 2 , 3.5 ) , ( 'gamma' , 
3 , 8.1 ) ] ,
    dtype = [ ( 'NAMES' , '|S5' ) , ( 'NR' , '<i2' ) , ( 'VAL' , '<f4' ) ] )
any ideas?

At this point the nicest way I know to do this is as follows:
    names = list( oldarray.dtype.names ) # get the record names
    names.append( 'VAL' ) # add the name for the new column
    olddescr = oldarray.dtype.descr # get the old type descriptor
    formats = [ olddescr[0][1] ] # initialize formats list, enter first 
format
    arrays = [ oldarray[ olddescr[0][0] ] ] # initialize arrays list, 
enter first array
    for i in range( 1 , len(olddescr) ) : # loop over remaining columns
        formats.append( olddescr[i][1] ) # add format of column
        arrays.append( oldarray[ olddescr[i][0] ] ) # add array of column
    formats.append( '<f4' ) # add format of new column
    arrays.append( array( [3.1,3.5,8.1] ) ) # add array of new column
    newarray = fromarrays( arrays , names = names , formats = formats ) 
# make new record array
but that is very ugly programming.
Any Ideas?

Thank you - MGB




From robert.kern at gmail.com  Mon Oct 13 20:41:24 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 13 Oct 2008 19:41:24 -0500
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <48F3E642.4040001@cornell.edu>
References: <48F3E642.4040001@cornell.edu>
Message-ID: <3d375d730810131741w1643b60br88623f418c8305d3@mail.gmail.com>

On Mon, Oct 13, 2008 at 19:22, Marc Berthoud <mgb11 at cornell.edu> wrote:
> What is the best way to add a column to a record array?
>
> Example:
> Add to this array
>  recarray( [ ( 'alpha' , 1 ) , ( 'beta ' , 2 ) , ( 'gamma' , 3 ) ] ,
>    dtype = [ ( 'NAMES' , '|S5' ) , ( 'NR' , '>i2' ) ] )
> the following column
>   array( [ 3.1 ,  3.5 ,  8.1 ] )
> to make that array
>  recarray( [ ( 'alpha' , 1, 3.1 ), ( 'beta ' , 2 , 3.5 ) , ( 'gamma' ,
> 3 , 8.1 ) ] ,
>    dtype = [ ( 'NAMES' , '|S5' ) , ( 'NR' , '<i2' ) , ( 'VAL' , '<f4' ) ] )
> any ideas?
>
> At this point the nicest way I know to do this is as follows:
>    names = list( oldarray.dtype.names ) # get the record names
>    names.append( 'VAL' ) # add the name for the new column
>    olddescr = oldarray.dtype.descr # get the old type descriptor
>    formats = [ olddescr[0][1] ] # initialize formats list, enter first
> format
>    arrays = [ oldarray[ olddescr[0][0] ] ] # initialize arrays list,
> enter first array
>    for i in range( 1 , len(olddescr) ) : # loop over remaining columns
>        formats.append( olddescr[i][1] ) # add format of column
>        arrays.append( oldarray[ olddescr[i][0] ] ) # add array of column
>    formats.append( '<f4' ) # add format of new column
>    arrays.append( array( [3.1,3.5,8.1] ) ) # add array of new column
>    newarray = fromarrays( arrays , names = names , formats = formats )
> # make new record array
> but that is very ugly programming.
> Any Ideas?

This is somewhat more straightforward:

http://projects.scipy.org/pipermail/numpy-discussion/2007-September/029357.html

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From wnbell at gmail.com  Mon Oct 13 22:57:15 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Mon, 13 Oct 2008 22:57:15 -0400
Subject: [SciPy-user] scipy.sparse: coo_matrix ignores
	sum_duplicates=False
In-Reply-To: <2b1c8c4f0810131533x430b6254x140691ce2a934942@mail.gmail.com>
References: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>
	<d05265cb0810130735p18d8c237ybe8fa54163a35317@mail.gmail.com>
	<2b1c8c4f0810130752w138837fan9ef20c59204a1eb2@mail.gmail.com>
	<d05265cb0810131230l367b30c2ha4a9b0815144a64@mail.gmail.com>
	<2b1c8c4f0810131533x430b6254x140691ce2a934942@mail.gmail.com>
Message-ID: <d05265cb0810131957jf08effag3fd2ad72795299eb@mail.gmail.com>

On Mon, Oct 13, 2008 at 6:33 PM, James Philbin <philbinj at gmail.com> wrote:
>
> same. I think dok_matrix is fine for my needs. BTW, i've found that
> __setitem__ is v slow for dok_matrix. Is this just because of the
> checks which are made? Using dict.__setitem__(mat, (r,c), val) is
> about an order of magnitude faster.

I don't use dok_matrix, so I don't know why it would be that much
slower.  If you can speed it up and submit a patch I'd happily apply
it.

> I'm not arguing that summing duplicate entries is not desirable. I'm
> just arguing that a function which reads .tocsr(sum_duplicates=False)
> and then sums the duplicates implicitly is misnamed.

Please understand, it *does not* sum the duplicates.  As I illustrated
before, the duplicates are carried over to the CSR format.  It's just
that CSR->dense *does* sum duplicates.

I agree that sum_duplicates=False is somewhat ambiguous, do you have a
suggestion for how this could be made more clear?  For instance, would
an interface like:
  coo_matrix.tocsr(duplicates='sum')
  coo_matrix.tocsr(duplicates='last')
  coo_matrix.tocsr(duplicates='max')
be preferred?  If I understand correctly, you'd want to use
.tocsr(duplicates='last').

Another question is whether we want to put this in the COO->CSR (and
CSC) conversions.  At this point, I think COO->CSR should *always* sum
duplicates together and we should instead provide a separate function
or member function of coo_matrix that provides additional options,
like 'last', 'max', etc.  In general, any binary operator (T,T) -> T
could be used as an accumulator, but we would provide the most common
options.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From kwmsmith at gmail.com  Mon Oct 13 23:07:35 2008
From: kwmsmith at gmail.com (Kurt Smith)
Date: Mon, 13 Oct 2008 22:07:35 -0500
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <ce557a360810130825q1fbad70i7fbed40f868daa16@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<ce557a360810130825q1fbad70i7fbed40f868daa16@mail.gmail.com>
Message-ID: <c266a7930810132007i36ffdb23n2e3ec83a6c5a6bf2@mail.gmail.com>

On Mon, Oct 13, 2008 at 10:25 AM, Anne Archibald
<peridot.faceted at gmail.com> wrote:
> 2008/10/12 Kurt Smith <kwmsmith at gmail.com>:
>
>> I've recently acquired a new Ubuntu system, and unfortunately some 3rd
>> party software I'd like to use won't work under the 64 bit version of
>> the OS, yet.  It's not a deal breaker, however, and I'm willing to
>> work around it as long as there are compelling reasons to use 64 bit
>> for numerical work, specifically under scipy, etc.  I'm not well
>> versed in the low-level pros and cons of 64 bit vs 32 bit, esp. as to
>> how it affects scipy performance.  Can you enlighten me?  I"m on an
>> Intel Core 2 Duo, 3.16 GHz, E8500, 4 GB RAM.  Will using 64 bit yield
>> great benefits running scipy/numpy code automatically, or will it
>> require fine-tuning on my part?  Will 64 bit yield benefits besides
>> having a larger address space (besides the fact that my 4 GB RAM is
>> still addressable by 32 bits, anyway)?
>
> I think memory is the only major reason to prefer 64 bits. The speed
> difference will probably be minimal. But having a 64-bit address space
> means that if your applications get too big for RAM they just slow
> down, as opposed to crashing and forcing you to rewrite them. You will
> also find, I think, that running in 32-bit mode you cannot access the
> last gigabyte or so of physical RAM. This can make a big difference
> too.

Thanks to all for your answers; since posting I've worked out some
more issues and 64 bit looks better and better.  Indeed, it is at
least 2x "cooler."  ;-)  I'll do more benchmarks of my own, but so far
things look pretty good.

Kurt


From dcday137 at gmail.com  Mon Oct 13 23:36:16 2008
From: dcday137 at gmail.com (Collin Day)
Date: Mon, 13 Oct 2008 21:36:16 -0600
Subject: [SciPy-user] Difference between ffts?
Message-ID: <20081013213616.667c6106@Krypton.homenet>

Hi all,

I have looked around, but can't seem to find an answer.  I have been
trying the following (according to the Getting started page -
http://www.scipy.org/Getting_Started)

from scipy import *
a=zeros(1000)
a[:100]=1
b=fft(a)

plot(abs(b))

and I get what you would expect - the abs. value of a sinc function

rect(x) ->F-> sinc(Frequency)

now, if I try the Scipy.fftpack

import scipy.fftapck as S

c=S.fft(a)

figure()

plot(abs(c))

I get something I would expect if I did a FFT on a sine function (kind
of like dual spikes equally spaced)

I do see there is a difference in the packing or how the fft output is
represented.  How do you plot the data from fftpack so that it looks
correct?  Should I even bother?

Thanks for any help!

-Collin


From wnbell at gmail.com  Mon Oct 13 23:47:15 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Mon, 13 Oct 2008 23:47:15 -0400
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <c266a7930810132007i36ffdb23n2e3ec83a6c5a6bf2@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<ce557a360810130825q1fbad70i7fbed40f868daa16@mail.gmail.com>
	<c266a7930810132007i36ffdb23n2e3ec83a6c5a6bf2@mail.gmail.com>
Message-ID: <d05265cb0810132047v206f5fa8y20e6c212459cf2f9@mail.gmail.com>

On Mon, Oct 13, 2008 at 11:07 PM, Kurt Smith <kwmsmith at gmail.com> wrote:
>
> Thanks to all for your answers; since posting I've worked out some
> more issues and 64 bit looks better and better.  Indeed, it is at
> least 2x "cooler."  ;-)  I'll do more benchmarks of my own, but so far
> things look pretty good.
>

Good to hear!  Let us know what you find.  I'm sure there are others
in the same situation that would like to know what to expect.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From robert.kern at gmail.com  Mon Oct 13 23:48:41 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 13 Oct 2008 22:48:41 -0500
Subject: [SciPy-user] Difference between ffts?
In-Reply-To: <20081013213616.667c6106@Krypton.homenet>
References: <20081013213616.667c6106@Krypton.homenet>
Message-ID: <3d375d730810132048k3c14eaa8oa71f9dc9b12772c1@mail.gmail.com>

On Mon, Oct 13, 2008 at 22:36, Collin Day <dcday137 at gmail.com> wrote:
> Hi all,
>
> I have looked around, but can't seem to find an answer.  I have been
> trying the following (according to the Getting started page -
> http://www.scipy.org/Getting_Started)
>
> from scipy import *
> a=zeros(1000)
> a[:100]=1
> b=fft(a)
>
> plot(abs(b))
>
> and I get what you would expect - the abs. value of a sinc function
>
> rect(x) ->F-> sinc(Frequency)
>
> now, if I try the Scipy.fftpack
>
> import scipy.fftapck as S
>
> c=S.fft(a)
>
> figure()
>
> plot(abs(c))
>
> I get something I would expect if I did a FFT on a sine function (kind
> of like dual spikes equally spaced)

Can you show us the plots? I don't see a difference.

> I do see there is a difference in the packing or how the fft output is
> represented.  How do you plot the data from fftpack so that it looks
> correct?  Should I even bother?

Typically, I will use fftfreq() to get the "X" values in packed form
and then use fftshift() on both X and Y to "unpack" the arrays.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From philbinj at gmail.com  Tue Oct 14 04:56:42 2008
From: philbinj at gmail.com (James Philbin)
Date: Tue, 14 Oct 2008 09:56:42 +0100
Subject: [SciPy-user] scipy.sparse: coo_matrix ignores
	sum_duplicates=False
In-Reply-To: <d05265cb0810131957jf08effag3fd2ad72795299eb@mail.gmail.com>
References: <2b1c8c4f0810130704n7fb317bcr1ee613c1a5975e29@mail.gmail.com>
	<d05265cb0810130735p18d8c237ybe8fa54163a35317@mail.gmail.com>
	<2b1c8c4f0810130752w138837fan9ef20c59204a1eb2@mail.gmail.com>
	<d05265cb0810131230l367b30c2ha4a9b0815144a64@mail.gmail.com>
	<2b1c8c4f0810131533x430b6254x140691ce2a934942@mail.gmail.com>
	<d05265cb0810131957jf08effag3fd2ad72795299eb@mail.gmail.com>
Message-ID: <2b1c8c4f0810140156n64a963f8v1ccc4d8dfdfc74c3@mail.gmail.com>

> Please understand, it *does not* sum the duplicates.  As I illustrated
> before, the duplicates are carried over to the CSR format.  It's just
> that CSR->dense *does* sum duplicates.
No, I do understand. This is why I said 'implicitly'. The CSR keeps
the duplicates, but then always behaves as if they'd been summed. To
the user, therefore, the effect is the same.

> I agree that sum_duplicates=False is somewhat ambiguous, do you have a
> suggestion for how this could be made more clear?  For instance, would
> an interface like:
>  coo_matrix.tocsr(duplicates='sum')
>  coo_matrix.tocsr(duplicates='last')
>  coo_matrix.tocsr(duplicates='max')
> be preferred?  If I understand correctly, you'd want to use
> .tocsr(duplicates='last').
I'm not sure it's worth you having to implement something which i'm
not sure that many people really need. I don't want scipy.sparse to
get feature-itis. I'd be happy if the sum_duplicates parameter was
removed altogether, with the standard behaviour being the one for
sum_duplicates=True. Then just state clearly in the docstring what
that behaviour is.

> Another question is whether we want to put this in the COO->CSR (and
> CSC) conversions.  At this point, I think COO->CSR should *always* sum
> duplicates together and we should instead provide a separate function
> or member function of coo_matrix that provides additional options,
> like 'last', 'max', etc.  In general, any binary operator (T,T) -> T
> could be used as an accumulator, but we would provide the most common
> options.
This seems fine, but I don't in general like modal options as they
tend to be bug-prone. Maybe a separate member of coo_matrix called
'merge_duplicates' which would apply some operation in-place on
coo_matrix where the user could specify 'sum', 'max', 'first', 'last',
etc.

Thanks,
James


From jdh2358 at gmail.com  Tue Oct 14 06:35:06 2008
From: jdh2358 at gmail.com (John Hunter)
Date: Tue, 14 Oct 2008 05:35:06 -0500
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <3d375d730810131741w1643b60br88623f418c8305d3@mail.gmail.com>
References: <48F3E642.4040001@cornell.edu>
	<3d375d730810131741w1643b60br88623f418c8305d3@mail.gmail.com>
Message-ID: <88e473830810140335y7686d87m9254587f55aaac84@mail.gmail.com>

On Mon, Oct 13, 2008 at 7:41 PM, Robert Kern <robert.kern at gmail.com> wrote:

> This is somewhat more straightforward:
>
> http://projects.scipy.org/pipermail/numpy-discussion/2007-September/029357.html

I took Robert's suggestion from the link above and added
rec_append_fields to matplotlib.mlab  -- I think it may have been
called rec_append_field in 0.98.3, but we altered it in svn HEAD to
support multiple column adds.  There are a number of nice helper
functions for recarrays there

   * rec2txt          : pretty print a record array
   * rec2csv          : store record array in CSV file
   * csv2rec          : import record array from CSV file with type inspection
   * rec_append_fields: adds  field(s)/array(s) to record array
   * rec_drop_fields  : drop fields from record array
   * rec_join         : join two record arrays on sequence of fields
   * rec_groupby      : summarize data by groups (similar to SQL GROUP BY)
   * rec_summarize    : helper code to filter rec array fields into new fields

rec_join is really nice -- supports inner and outer joins with default
fill values and customizable postfixing of column names when joining
two record arrays with identically named fields.

Here is an example showing many of these functions in action

"""
Illustrate the rec array utility funcitons by loading prices from a
csv file, computing the daily returns, appending the results to the
record arrays, joining on date
"""
import urllib
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.mlab as mlab

# grab the price data off yahoo
u1 = urllib.urlretrieve('http://ichart.finance.yahoo.com/table.csv?s=AAPL&d=9&e=14&f=2008&g=d&a=8&b=7&c=1984&ignore=.csv')
u2 = urllib.urlretrieve('http://ichart.finance.yahoo.com/table.csv?s=GOOG&d=9&e=14&f=2008&g=d&a=8&b=7&c=1984&ignore=.csv')

# load the CSV files into record arrays
r1 = mlab.csv2rec(file(u1[0]))
r2 = mlab.csv2rec(file(u2[0]))

# compute the daily returns and add these columns to the arrays
gains1 = np.zeros_like(r1.adj_close)
gains2 = np.zeros_like(r2.adj_close)
gains1[1:] = np.diff(r1.adj_close)/r1.adj_close[:-1]
gains2[1:] = np.diff(r2.adj_close)/r2.adj_close[:-1]
r1 = mlab.rec_append_fields(r1, 'gains', gains1)
r2 = mlab.rec_append_fields(r2, 'gains', gains2)

# now join them by date; the default postfixes are 1 and 2
r = mlab.rec_join('date', r1, r2)

# long appl, short goog
g = r.gains1-r.gains2
tr = (1+g).cumprod()  # the total return

# plot the return
fig = plt.figure()
ax = fig.add_subplot(111)
ax.plot(r.date, tr)
ax.set_title('total return: long appl, short goog')
ax.grid()
fig.autofmt_xdate()
plt.show()


From hwchen.mailman at gmail.com  Tue Oct 14 11:02:09 2008
From: hwchen.mailman at gmail.com (Huang-Wen Chen)
Date: Tue, 14 Oct 2008 11:02:09 -0400
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <88e473830810140335y7686d87m9254587f55aaac84@mail.gmail.com>
References: <48F3E642.4040001@cornell.edu>
	<3d375d730810131741w1643b60br88623f418c8305d3@mail.gmail.com>
	<88e473830810140335y7686d87m9254587f55aaac84@mail.gmail.com>
Message-ID: <368e8c230810140802x4a973009l749c1da77a79c57d@mail.gmail.com>

On Tue, Oct 14, 2008 at 6:35 AM, John Hunter <jdh2358 at gmail.com> wrote:

> I took Robert's suggestion from the link above and added
> rec_append_fields to matplotlib.mlab  -- I think it may have been
> called rec_append_field in 0.98.3, but we altered it in svn HEAD to
> support multiple column adds.  There are a number of nice helper
> functions for recarrays there
>
> Here is an example showing many of these functions in action


The example is amazing. It's so simple to do the elegant job, but I think it
requires some modification because the the size of the two dataset is
different:

In [68]: r1.shape
Out[68]: (6080,)

In [69]: r2.shape
Out[69]: (1046,)

Maybe we need to do some trimming in the example such as:

l = min(r1.shape[0], r2.shape[0])
r1,r2=r1[:l],r2[:l]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081014/9aab3996/attachment.html>

From fredmfp at gmail.com  Tue Oct 14 11:40:59 2008
From: fredmfp at gmail.com (fred)
Date: Tue, 14 Oct 2008 17:40:59 +0200
Subject: [SciPy-user] 64 bit and 32 bit on Ubuntu
In-Reply-To: <ce557a360810130825q1fbad70i7fbed40f868daa16@mail.gmail.com>
References: <c266a7930810121145p4df42cban7c7c044d2a1e041d@mail.gmail.com>
	<ce557a360810130825q1fbad70i7fbed40f868daa16@mail.gmail.com>
Message-ID: <48F4BD8B.8070408@gmail.com>

Anne Archibald a ?crit :

> I think memory is the only major reason to prefer 64 bits. The speed
> difference will probably be minimal. But having a 64-bit address space
> means that if your applications get too big for RAM they just slow
> down, as opposed to crashing and forcing you to rewrite them. You will
> also find, I think, that running in 32-bit mode you cannot access the
> last gigabyte or so of physical RAM. This can make a big difference
> too.

I have computed interpolation on 3D data (50x50x50) using kriging method.

On the same computer (Core2 Duo), the same OS (debian lenny), same 
configuration (same packages), same compiler (f2py with ifort 10.1), 
same options (-xT on both arch), I get:

on i686 arch : 3358 seconds

on x86_64 arch : 2130 seconds

I don't say this is relevant, or whatsoever.
I only say: that's it ;-)

However, I would be happy to hear some comment/objection about this result.


Only my 2 cts.

Cheers,

-- 
Fred


From jdh2358 at gmail.com  Tue Oct 14 11:53:08 2008
From: jdh2358 at gmail.com (John Hunter)
Date: Tue, 14 Oct 2008 10:53:08 -0500
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <368e8c230810140802x4a973009l749c1da77a79c57d@mail.gmail.com>
References: <48F3E642.4040001@cornell.edu>
	<3d375d730810131741w1643b60br88623f418c8305d3@mail.gmail.com>
	<88e473830810140335y7686d87m9254587f55aaac84@mail.gmail.com>
	<368e8c230810140802x4a973009l749c1da77a79c57d@mail.gmail.com>
Message-ID: <88e473830810140853o5e6d5e99g5cbf04a0989febf5@mail.gmail.com>

On Tue, Oct 14, 2008 at 10:02 AM, Huang-Wen Chen
<hwchen.mailman at gmail.com> wrote:

> The example is amazing. It's so simple to do the elegant job, but I think it
> requires some modification because the the size of the two dataset is
> different:
>
> In [68]: r1.shape
> Out[68]: (6080,)
>
> In [69]: r2.shape
> Out[69]: (1046,)

No, that is one of the main points of the example in the call to
rec_join -- it does an inner join (intersection)  aligned by date.
Since it is an inner join, dates in one that are not in the other are
dropped.  It can also do an outer join (union) using the "jointype"
keyword arg to rec_join.

But thanks for the kind words on the example -- I agree.  Record
arrays are really powerful data structures and with some of the
functions in mlab, which hopefully will end up in some form in numpy
eventually, you have much of the flexibility of sql tables.  See for
example 'group by' using
record arrays in rec_groupby_demo.py at

  http://matplotlib.svn.sourceforge.net/viewvc/matplotlib/trunk/matplotlib/examples/misc/

JDH


From cmac at mit.edu  Tue Oct 14 12:19:21 2008
From: cmac at mit.edu (Christopher MacMinn)
Date: Tue, 14 Oct 2008 12:19:21 -0400
Subject: [SciPy-user] ImportError: cannot import name Tester
Message-ID: <05AECBC4-87D0-470E-A9C6-5FE9DB851E0C@mit.edu>

Hi folks -

I just installed Scipy 0.7.0 build 4797, and import is failing:

# ---------------------------------------------------------

In [3]: import scipy
---------------------------------------------------------------------------
ImportError                               Traceback (most recent call  
last)

/Library/Python/2.5/site-packages/<ipython console> in <module>()

/Library/Python/2.5/site-packages/scipy/__init__.py in <module>()
     86     __doc__ += pkgload.get_pkgdocs()
     87
---> 88 from numpy.testing import Tester
     89 test = Tester().test
     90 bench = Tester().bench

ImportError: cannot import name Tester

In [4]:

# ---------------------------------------------------------

I imagine I can fix this by commenting out the troublesome lines in  
__init__.py, but perhaps it is a bug in this build?  I did not have  
this problem with build 4786.

I'm running Python 2.5 and numpy 1.2.0 on Mac OS 10.5.

Best, Chris MacMinn


From robert.kern at gmail.com  Tue Oct 14 12:23:17 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 14 Oct 2008 11:23:17 -0500
Subject: [SciPy-user] ImportError: cannot import name Tester
In-Reply-To: <05AECBC4-87D0-470E-A9C6-5FE9DB851E0C@mit.edu>
References: <05AECBC4-87D0-470E-A9C6-5FE9DB851E0C@mit.edu>
Message-ID: <3d375d730810140923k4f715d15o6fab2d0f419b0af3@mail.gmail.com>

On Tue, Oct 14, 2008 at 11:19, Christopher MacMinn <cmac at mit.edu> wrote:
> Hi folks -
>
> I just installed Scipy 0.7.0 build 4797, and import is failing:
>
> # ---------------------------------------------------------
>
> In [3]: import scipy
> ---------------------------------------------------------------------------
> ImportError                               Traceback (most recent call
> last)
>
> /Library/Python/2.5/site-packages/<ipython console> in <module>()
>
> /Library/Python/2.5/site-packages/scipy/__init__.py in <module>()
>     86     __doc__ += pkgload.get_pkgdocs()
>     87
> ---> 88 from numpy.testing import Tester
>     89 test = Tester().test
>     90 bench = Tester().bench
>
> ImportError: cannot import name Tester
>
> In [4]:
>
> # ---------------------------------------------------------
>
> I imagine I can fix this by commenting out the troublesome lines in
> __init__.py, but perhaps it is a bug in this build?  I did not have
> this problem with build 4786.
>
> I'm running Python 2.5 and numpy 1.2.0 on Mac OS 10.5.

Hmm, numpy 1.2.0 should have numpy.testing.Tester. Can you
double-check that that is the numpy which is actually getting picked
up?

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From pgmdevlist at gmail.com  Tue Oct 14 12:28:42 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 14 Oct 2008 12:28:42 -0400
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <88e473830810140853o5e6d5e99g5cbf04a0989febf5@mail.gmail.com>
References: <48F3E642.4040001@cornell.edu>
	<368e8c230810140802x4a973009l749c1da77a79c57d@mail.gmail.com>
	<88e473830810140853o5e6d5e99g5cbf04a0989febf5@mail.gmail.com>
Message-ID: <200810141228.42850.pgmdevlist@gmail.com>

John,
Do you plan to have your modifications part of numpy.records ? In any case, 
I'll try to check whether it is easy to add support to missing data: 
MaskedArrays should now support with flexible-types.


From cmac at mit.edu  Tue Oct 14 12:32:44 2008
From: cmac at mit.edu (Christopher MacMinn)
Date: Tue, 14 Oct 2008 12:32:44 -0400
Subject: [SciPy-user] SciPy-user Digest, Vol 62, Issue 27
In-Reply-To: <mailman.819.1224001403.2878.scipy-user@scipy.org>
References: <mailman.819.1224001403.2878.scipy-user@scipy.org>
Message-ID: <A26E1F73-1F15-4F9F-98BD-EF29A772CF10@mit.edu>

>> I just installed Scipy 0.7.0 build 4797, and import is failing:
>>
>>
>>
>> In [3]: import scipy
>> ...

>> ImportError: cannot import name Tester
>>
>>
>> I'm running Python 2.5 and numpy 1.2.0 on Mac OS 10.5.
>
>
> Hmm, numpy 1.2.0 should have numpy.testing.Tester. Can you
> double-check that that is the numpy which is actually getting picked
> up?


I just fired up ipython again and imported numpy and scipy with no  
problems.  It must have been a path issue or something that cleared  
itself up, although that strikes me as strange.  In any case, I'm good  
now.

Sorry for the false alarm, and thanks!

Best, Chris


From jdh2358 at gmail.com  Tue Oct 14 12:55:25 2008
From: jdh2358 at gmail.com (John Hunter)
Date: Tue, 14 Oct 2008 11:55:25 -0500
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <200810141228.42850.pgmdevlist@gmail.com>
References: <48F3E642.4040001@cornell.edu>
	<368e8c230810140802x4a973009l749c1da77a79c57d@mail.gmail.com>
	<88e473830810140853o5e6d5e99g5cbf04a0989febf5@mail.gmail.com>
	<200810141228.42850.pgmdevlist@gmail.com>
Message-ID: <88e473830810140955u6f5d830cueb24a3f4bc980458@mail.gmail.com>

On Tue, Oct 14, 2008 at 11:28 AM, Pierre GM <pgmdevlist at gmail.com> wrote:
> John,
> Do you plan to have your modifications part of numpy.records ? In any case,
> I'll try to check whether it is easy to add support to missing data:
> MaskedArrays should now support with flexible-types.

I do not have concrete plans, but I have spoken with Jarrod about
moving some of these over, making some of them record array methods,
others available in the np.rec namespace.  I think the consensus is
that these are useful and belong in numpy, but we are awaiting someone
to do the port.

On the subject of masked record arrays.  We added masked array support
to mlab.csv2rec some time ago and it has caused no shortage of
headaches because of differences in the interface for objects for
masked record arrays and regular recarrays.

The following example shows a record array with a 'date' column which
is a O4 python object type. Here is the behavior of the recarray

  In [212]: !cat test1.csv
  date,age,name
  2008-01-01,10,'tom'
  2008-01-02,11,'dick'
  2008-01-03,12,'harry'
  In [213]: r1 = mlab.csv2rec('test1.csv')

  In [214]: type(r1)
  Out[214]: <class 'numpy.core.records.recarray'>

  In [215]: r1.dtype
  Out[215]: dtype([('date', '|O4'), ('age', '<i4'), ('name', '|S7')])

  In [216]: print r1[0].date.year
  2008

In particular, on a given row of the recarray, I can call object
methods and access object attributes.

In the next example, the data file has a missing value on the last row
in the 'age' column, so we return a masked record array

  In [217]: !cat test2.csv
  date,age,name
  2008-01-01,10,'tom'
  2008-01-02,11,'dick'
  2008-01-03,,'harry'
  In [218]: type(r2)
  Out[218]: <class 'numpy.ma.mrecords.MaskedRecords'>

  In [219]: print r2.dtype
  [('date', '|O4'), ('age', '<i4'), ('name', '|S7')]

  In [220]: r2[0].date.year
  ------------------------------------------------------------
  Traceback (most recent call last):
    File "<ipython console>", line 1, in ?
  AttributeError: 'MaskedArray' object has no attribute 'year'

It would help us a lot in this regard if we could access the
underlying object.  Is there a reason why the masked array behaves
differently when it comes to accessing the underlying object methods
and is there a sensible way to make them compatible?

Thanks,
JDH


From pgmdevlist at gmail.com  Tue Oct 14 13:14:22 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Tue, 14 Oct 2008 13:14:22 -0400
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <88e473830810140955u6f5d830cueb24a3f4bc980458@mail.gmail.com>
References: <48F3E642.4040001@cornell.edu>
	<200810141228.42850.pgmdevlist@gmail.com>
	<88e473830810140955u6f5d830cueb24a3f4bc980458@mail.gmail.com>
Message-ID: <200810141314.22369.pgmdevlist@gmail.com>

John,
MaskedRecords have always been the poor lost child of MaskedArrays. In 1.3.x 
and 1.2.1, I tried to improve the support of flexible-type, and MaskedRecords 
should follow. Of course, I can only check whether it works in the cases I 
need.

Could you send me a self-contained example showing what you expect (example + 
class definitions...) ?
My guess here is that when you query the field date on a standard 'recarray', 
you get a date object, which has a 'year' attribute. However, when you query 
it as a MaskedRecords, the 'date' column is transformed into a MaskedArray 
(in order to handle missing data), and then you can't directly access 
the 'year' attribute: you'd need to access 'date._data.year' instead.

Once again, that's a very first guess, I'd need to do some testing.
On a side note, TimeSeries objects support flexible type as well, if you catch 
my drift...


P.


> In the next example, the data file has a missing value on the last row
> in the 'age' column, so we return a masked record array
>
>   In [217]: !cat test2.csv
>   date,age,name
>   2008-01-01,10,'tom'
>   2008-01-02,11,'dick'
>   2008-01-03,,'harry'
>   In [218]: type(r2)
>   Out[218]: <class 'numpy.ma.mrecords.MaskedRecords'>
>
>   In [219]: print r2.dtype
>   [('date', '|O4'), ('age', '<i4'), ('name', '|S7')]
>
>   In [220]: r2[0].date.year
>   ------------------------------------------------------------
>   Traceback (most recent call last):
>     File "<ipython console>", line 1, in ?
>   AttributeError: 'MaskedArray' object has no attribute 'year'
>
> It would help us a lot in this regard if we could access the
> underlying object.  Is there a reason why the masked array behaves
> differently when it comes to accessing the underlying object methods
> and is there a sensible way to make them compatible?
>
> Thanks,
> JDH
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user




From oliphant at enthought.com  Tue Oct 14 18:55:44 2008
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Tue, 14 Oct 2008 17:55:44 -0500
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <88e473830810140853o5e6d5e99g5cbf04a0989febf5@mail.gmail.com>
References: <48F3E642.4040001@cornell.edu>	<3d375d730810131741w1643b60br88623f418c8305d3@mail.gmail.com>	<88e473830810140335y7686d87m9254587f55aaac84@mail.gmail.com>	<368e8c230810140802x4a973009l749c1da77a79c57d@mail.gmail.com>
	<88e473830810140853o5e6d5e99g5cbf04a0989febf5@mail.gmail.com>
Message-ID: <48F52370.80609@enthought.com>

John Hunter wrote:
>
> No, that is one of the main points of the example in the call to
> rec_join -- it does an inner join (intersection)  aligned by date.
> Since it is an inner join, dates in one that are not in the other are
> dropped.  It can also do an outer join (union) using the "jointype"
> keyword arg to rec_join.
>
> But thanks for the kind words on the example -- I agree.  Record
> arrays are really powerful data structures and with some of the
> functions in mlab, which hopefully will end up in some form in numpy
>   
I hope so too.  These are good examples, John.  Thanks for the continual 
high-quality code.

-Travis



From dcday137 at gmail.com  Tue Oct 14 19:46:22 2008
From: dcday137 at gmail.com (Collin Day)
Date: Tue, 14 Oct 2008 17:46:22 -0600
Subject: [SciPy-user] Difference between ffts? - plots of what i am getting.
Message-ID: <20081014174622.128f368b@Krypton.homenet>

No problem.  Sorry for not replying directly - I get the digests so my
mail box remains uncluttered.  Anyway,  I have attached two files.
Following are the lines of code I used:


import scipy as S
import scipy.fftpack as SF
a=zeros(1000)
b=S.fft(a)
plot(abs(b))

<S_fft_b.png attached>

c=SF.fft(a)
figure()
plot(c)

<SF_fft_c.png attached>

Hope this helps.  Thanks for looking at it!

-C
-------------- next part --------------
A non-text attachment was scrubbed...
Name: S_fft_b.png
Type: image/png
Size: 25934 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081014/4a25f735/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: SF_fft_c.png
Type: image/png
Size: 15114 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081014/4a25f735/attachment-0001.png>

From dcday137 at gmail.com  Tue Oct 14 19:51:34 2008
From: dcday137 at gmail.com (Collin Day)
Date: Tue, 14 Oct 2008 17:51:34 -0600
Subject: [SciPy-user] Correction Difference between ffts? - plots of what i
	am getting.
Message-ID: <20081014175134.3875433d@Krypton.homenet>

Oops - I meant plot(abs(c)) for the code on that last plot statement.

Thanks again.

-C


From robert.kern at gmail.com  Tue Oct 14 20:53:13 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 14 Oct 2008 19:53:13 -0500
Subject: [SciPy-user] Difference between ffts? - plots of what i am
	getting.
In-Reply-To: <20081014174622.128f368b@Krypton.homenet>
References: <20081014174622.128f368b@Krypton.homenet>
Message-ID: <3d375d730810141753y7ace1e0i34bad7eb4e055c60@mail.gmail.com>

On Tue, Oct 14, 2008 at 18:46, Collin Day <dcday137 at gmail.com> wrote:
> No problem.  Sorry for not replying directly - I get the digests so my
> mail box remains uncluttered.  Anyway,  I have attached two files.
> Following are the lines of code I used:

Hmm. Did you build scipy with any special FFT libraries?

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From david at ar.media.kyoto-u.ac.jp  Tue Oct 14 21:55:14 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 15 Oct 2008 10:55:14 +0900
Subject: [SciPy-user] Difference between ffts? - plots of what i am
	getting.
In-Reply-To: <20081014174622.128f368b@Krypton.homenet>
References: <20081014174622.128f368b@Krypton.homenet>
Message-ID: <48F54D82.5000005@ar.media.kyoto-u.ac.jp>

Collin Day wrote:
> No problem.  Sorry for not replying directly - I get the digests so my
> mail box remains uncluttered.  Anyway,  I have attached two files.
> Following are the lines of code I used:
>   

Hi Collin,

    Would it be possible to provide the exact script which produces the
graph ? One graph looks like a log graph, the other linear, and I can't
see how it would be possible for both graphs to be produced by the above
script, even assuming bugs in scipy ?

Also, to know which fft was used, could you execute the following
command and give us the result:

import scipy
scipy.show_config()

thanks,

David


From dcday137 at gmail.com  Tue Oct 14 23:38:30 2008
From: dcday137 at gmail.com (Collin Day)
Date: Wed, 15 Oct 2008 03:38:30 +0000 (UTC)
Subject: [SciPy-user]
	=?utf-8?q?Difference_between_ffts=3F_-_plots_of_what?=
	=?utf-8?q?_i_am=09getting=2E?=
References: <20081014174622.128f368b@Krypton.homenet>
	<48F54D82.5000005@ar.media.kyoto-u.ac.jp>
Message-ID: <loom.20081015T032936-16@post.gmane.org>

David Cournapeau <david <at> ar.media.kyoto-u.ac.jp> writes:

> 
> Collin Day wrote:
> > No problem.  Sorry for not replying directly - I get the digests so my
> > mail box remains uncluttered.  Anyway,  I have attached two files.
> > Following are the lines of code I used:
> >   
> 
> Hi Collin,
> 
>     Would it be possible to provide the exact script which produces the
> graph ? One graph looks like a log graph, the other linear, and I can't
> see how it would be possible for both graphs to be produced by the above
> script, even assuming bugs in scipy ?
> 
> Also, to know which fft was used, could you execute the following
> command and give us the result:
> 
> import scipy
> scipy.show_config()
> 
> thanks,
> 
> David
> 
Sure.  The following script will produce the plots:


#!/usr/bin/python

from pylab import *
import scipy as S
import scipy.fftpack as SF

a=zeros(1000)
a[:100]=1
b=S.fft(a)
c=SF.fft(a)
figure(1)
plot(abs(b))
show(1)
figure(2)
plot(abs(c))
show(2)

and scipy was compiled using the FFTW library.  The only thing I can thinkk of
would be to recompile not using it.

Hope that helps.  I really hope I have not done something stupid and missed it.
 Anyway, thanks again everyone!

-C





From dcday137 at gmail.com  Tue Oct 14 23:44:55 2008
From: dcday137 at gmail.com (Collin Day)
Date: Wed, 15 Oct 2008 03:44:55 +0000 (UTC)
Subject: [SciPy-user]
	=?utf-8?q?Difference_between_ffts=3F_-_plots_of_what?=
	=?utf-8?q?_i_am=09getting=2E?=
References: <20081014174622.128f368b@Krypton.homenet>
	<48F54D82.5000005@ar.media.kyoto-u.ac.jp>
Message-ID: <loom.20081015T034324-752@post.gmane.org>

David Cournapeau <david <at> ar.media.kyoto-u.ac.jp> writes:

> 
> Collin Day wrote:
> > No problem.  Sorry for not replying directly - I get the digests so my
> > mail box remains uncluttered.  Anyway,  I have attached two files.
> > Following are the lines of code I used:
> >   
> 
> Hi Collin,
> 
>     Would it be possible to provide the exact script which produces the
> graph ? One graph looks like a log graph, the other linear, and I can't
> see how it would be possible for both graphs to be produced by the above
> script, even assuming bugs in scipy ?
> 
> Also, to know which fft was used, could you execute the following
> command and give us the result:
> 
> import scipy
> scipy.show_config()
> 
> thanks,
> 
> David
> 

Sorry - forgot to post the following:

lapack_info:
    libraries = ['lapack']
    library_dirs = ['/usr/lib']
    language = f77

lapack_opt_info:
    libraries = ['lapack', 'blas']
    library_dirs = ['/usr/lib']
    language = f77
    define_macros = [('NO_ATLAS_INFO', 1)]

umfpack_info:
    libraries = ['umfpack', 'amd']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_UMFPACK_H', None), ('SCIPY_AMD_H', None)]
    swig_opts = ['-I/usr/include', '-I/usr/include']
    include_dirs = ['/usr/include']

blas_info:
    libraries = ['blas']
    library_dirs = ['/usr/lib']
    language = f77

atlas_threads_info:
  NOT AVAILABLE

fftw2_info:
    libraries = ['rfftw', 'fftw']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_FFTW_H', None)]
    include_dirs = ['/usr/include']

atlas_blas_info:
  NOT AVAILABLE

djbfft_info:
  NOT AVAILABLE

atlas_blas_threads_info:
  NOT AVAILABLE

fftw3_info:
  NOT AVAILABLE

blas_mkl_info:
  NOT AVAILABLE

amd_info:
    libraries = ['amd']
    library_dirs = ['/usr/lib']
    define_macros = [('SCIPY_AMD_H', None)]
    swig_opts = ['-I/usr/include']
    include_dirs = ['/usr/include']

blas_opt_info:
    libraries = ['blas']
    library_dirs = ['/usr/lib']
    language = f77
    define_macros = [('NO_ATLAS_INFO', 1)]

atlas_info:
  NOT AVAILABLE

lapack_mkl_info:
  NOT AVAILABLE

mkl_info:
  NOT AVAILABLE





From sahar at cmt.co.il  Wed Oct 15 04:49:32 2008
From: sahar at cmt.co.il (Sahar Vilan)
Date: Wed, 15 Oct 2008 10:49:32 +0200
Subject: [SciPy-user] saving raw image
In-Reply-To: <mailman.132.1224059690.2877.scipy-user@scipy.org>
Message-ID: <MNEGKHAFHCAFDOCPJMAMGEGHCBAA.sahar@cmt.co.il>

I used scipy for basic image processing of raw images, and I can't save
these images in the same format:
To open image I use:
	s = file(path, 'rb').read()
    	raw = fromstring(s, uint16).astype(int)

I try to save this matrix as raw image again but I get some garbage when I
open it in some viewer (Image-J, for instance).
Can anyone help me with this?

Thanks,
Sahar




*******************************************************************************************************
This e-mail message may contain confidential,and  privileged information or data that constitute proprietary information of CMT Medical Ltd. Any review or distribution by others is strictly prohibited. If you are not the intended recipient you are hereby notified that any use of this information or data by any other person is absolutely prohibited. If you are not the intended recipient, please delete all copies. Thank You. http://www.cmt.co.il
********************************************************************************************************





 
 
************************************************************************************
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals & computer viruses.
************************************************************************************





From zachary.pincus at yale.edu  Wed Oct 15 10:20:44 2008
From: zachary.pincus at yale.edu (Zachary Pincus)
Date: Wed, 15 Oct 2008 10:20:44 -0400
Subject: [SciPy-user] saving raw image
In-Reply-To: <MNEGKHAFHCAFDOCPJMAMGEGHCBAA.sahar@cmt.co.il>
References: <MNEGKHAFHCAFDOCPJMAMGEGHCBAA.sahar@cmt.co.il>
Message-ID: <68B3EEDF-FB62-40B5-944A-D85B4E28A0B5@yale.edu>

Hi Sahar,

Please send a minimal example of loading and saving an array that  
produces "garbage" output, if you could... (If you could send a small  
input file that would be helpful too.)

Also, does the garbage you see in ImageJ have some structure -- does  
it look streaky, like you can see rows of pixels that should be  
together, but the rows don't line up right?

One possible problem is that the array you are saving has been  
promoted to a different dtype (e.g. by participating in signed/float  
arithmetic), so the pixels you save are no longer uint16s. Unless you  
have an 'astype(uint16)' in your save code, this could be the issue,  
since the loading code you showed does immediately convert the array  
from uint16s. (This is why sending a complete example of the failure  
is useful...)

Another possible problem has to do with the order the pixels are read/ 
written out. Typically, images on disk are stored as rows of pixels  
next to one another -- this is "column major" or "fortran" order  
(going from one memory location to the next typically increments the x- 
value, except at row boundaries where the y-value is incremented, so  
it is said that the x-value "varies the fastest"). Typically, numpy  
arrays are created in row major or "C" order, where the y-value varies  
the fastest. When loading and manipulating images, you need either to  
make sure that the images are loaded in fortran-order, or reverse the  
x/y coordinates and shape.

So, e.g. when loading a 200x300 image:


s = file(path, 'rb').read()
raw = fromstring(s, uint16).astype(int)
image = raw.reshape((200,300), order='F')

now, image[30,40] gives the same pixel as coordinate (30,40) in ImageJ.

If you did:
image = raw.reshape((300,200), order='C')
then image[30,40] would give the same pixel as coordinate (40,30) in  
ImageJ.

Note that the 'C' order is default. So:
image = raw.reshape((200,300))
will give garbage, with the rows of pixels broken up along the wrong  
boundaries, giving rise to the "streaky" images I mentioned.

Finally, the tostring() method also takes an order option, so for  
saving images, you need:
raw = image.tostring(order='F')
(assuming that you reshaped the image as order 'F')

Zach Pincus



On Oct 15, 2008, at 4:49 AM, Sahar Vilan wrote:

> I used scipy for basic image processing of raw images, and I can't  
> save
> these images in the same format:
> To open image I use:
> 	s = file(path, 'rb').read()
>    	raw = fromstring(s, uint16).astype(int)
>
> I try to save this matrix as raw image again but I get some garbage  
> when I
> open it in some viewer (Image-J, for instance).
> Can anyone help me with this?
>
> Thanks,
> Sahar



From elcorto at gmx.net  Wed Oct 15 11:12:03 2008
From: elcorto at gmx.net (Steve Schmerler)
Date: Wed, 15 Oct 2008 17:12:03 +0200
Subject: [SciPy-user] npfile deprecated?
Message-ID: <20081015151203.GA7878@ramrod.starsheriffs.de>

Hi all

Since I'm currently in the need to store and read binary data a lot,
I choose npfile. Since [1] didn't mention it, I attempted to
update the doc here. I added scipy.io.npfile, numpy.load, numpy.save
and numpy.savez.  At work, I still use 0.6.0.  However, I found that in
the current svn HEAD, the scipy.io.npfile docstring says that it is
deprecated and I should use ndarray.tofile() and numpy.fromfile()
instead. But with the latter, I can't control byteorder, row- or
column-major order etc. That's a bit puzzling.   

I admitt that I haven't followed the development of the new numpy file
format very closely (since npfile worked nicely so far).  So, is the
.npy/.npz way the current standard to store platform-independent binary
arrays with all metainfo? 

[1] http://scipy.org/Cookbook/InputOutput

best,
steve


From scotta_2002 at yahoo.com  Wed Oct 15 12:28:16 2008
From: scotta_2002 at yahoo.com (Scott Askey)
Date: Wed, 15 Oct 2008 09:28:16 -0700 (PDT)
Subject: [SciPy-user] does scipy.signal.lti support mimo ?
Message-ID: <14420.43490.qm@web36504.mail.mud.yahoo.com>

Does scipy.signal.lti.ss2tf support multi-input/multi-ouput?

When I use ss2tf(A,B,C,D) it provides a proper transfer funtion if and only  if A is N*N, B is N*1 , C is 1*n and D is 1*1.  Is there any better documentation or a tutorial for the LTI suite?

V/R

Scott 


      


From wesmckinn at gmail.com  Wed Oct 15 13:45:20 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Wed, 15 Oct 2008 13:45:20 -0400
Subject: [SciPy-user] Matlab can interfere with Python?
Message-ID: <6c476c8a0810151045wddcef6fgab1b98d9b4f8578a@mail.gmail.com>

I have experienced this bizarre problem on two my coworker's machine who run
Matlab (7.6.0 rev 2008a) and I have been teaching to use Python/Numpy/Scipy.
Occasionally Matlab will somehow prevent Python from linking with various
DLLs or severely inhibit the execution of code. I am not really sure how
this could be (other than the Borg at the MathWorks not wanting anyone to
use open source software), but has anyone else experienced this? As soon as
I killed Matlab the problem disappeared.

Here is the linkage output from a Cython dll that was killed by this:

C:\MinGW\bin\gcc.exe -mno-cygwin -shared -s
build\temp.win32-2.5\Release\tseries.o
build\temp.win32-2.5\Release\tseries.def -LC:\Python25\libs -LC:\P
ython25\PCBuild -lpython25 -lmsvcr71 -o tseries.pyd

Maybe it's locking up the windows C runtime msvcr71?

Any help here would be appreciated.

Thanks,
Wes
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081015/39b67577/attachment.html>

From nwagner at iam.uni-stuttgart.de  Wed Oct 15 13:52:20 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 15 Oct 2008 19:52:20 +0200
Subject: [SciPy-user] linalg.lstsq
Message-ID: <web-112882343@uni-stuttgart.de>

Hi all,

I have a question concerning linalg.lstsq.

AFAIK, the residues of linalg.lstsq should be real in case 
of complex matrices.
Am I missing something ?


from scipy.linalg import lstsq
from scipy import rand
m = 9
n = 5
A = rand(m,n)+1j*rand(m,n)
b = rand(m)
x,residues,rank,s = lstsq(A,b)
print 'residues',residues # should be real

m = 9
n = 9
A = rand(m,n)+1j*rand(m,n)
b = rand(m)
x,residues,rank,s = lstsq(A,b)
print 'residues',residues # should be real order of 
machine precision


Nils


From 302302 at centrum.cz  Wed Oct 15 14:21:19 2008
From: 302302 at centrum.cz (302302)
Date: Wed, 15 Oct 2008 20:21:19 +0200
Subject: [SciPy-user] Time series graph
In-Reply-To: <200810152021.7089@centrum.cz>
References: <200810152017.29054@centrum.cz> <200810152018.13925@centrum.cz>
	<200810152019.29624@centrum.cz> <200810152020.8881@centrum.cz>
	<200810152021.7089@centrum.cz>
Message-ID: <200810152021.6780@centrum.cz>


Hi,

I need to plot a time series (the data in the database has around 10000 rows) and I need dynamically zoom and move with them.
Something like http://finance.google.com/finance?q=INDEXDJX:.DJI%20INDEXNASDAQ:.IXIC%20INDEXSP:.INX
Is there any efficient way how to do it with scipy or matplotlib?

Thanks Cenek



From robert.kern at gmail.com  Wed Oct 15 15:07:27 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 15 Oct 2008 14:07:27 -0500
Subject: [SciPy-user] npfile deprecated?
In-Reply-To: <20081015151203.GA7878@ramrod.starsheriffs.de>
References: <20081015151203.GA7878@ramrod.starsheriffs.de>
Message-ID: <3d375d730810151207i6efe0f6bwed94e65eeeae79ec@mail.gmail.com>

On Wed, Oct 15, 2008 at 10:12, Steve Schmerler <elcorto at gmx.net> wrote:
> Hi all
>
> Since I'm currently in the need to store and read binary data a lot,
> I choose npfile. Since [1] didn't mention it, I attempted to
> update the doc here. I added scipy.io.npfile, numpy.load, numpy.save
> and numpy.savez.  At work, I still use 0.6.0.  However, I found that in
> the current svn HEAD, the scipy.io.npfile docstring says that it is
> deprecated and I should use ndarray.tofile() and numpy.fromfile()
> instead. But with the latter, I can't control byteorder, row- or
> column-major order etc. That's a bit puzzling.
>
> I admitt that I haven't followed the development of the new numpy file
> format very closely (since npfile worked nicely so far).  So, is the
> .npy/.npz way the current standard to store platform-independent binary
> arrays with all metainfo?

Yup.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From pgmdevlist at gmail.com  Wed Oct 15 16:04:20 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Wed, 15 Oct 2008 16:04:20 -0400
Subject: [SciPy-user] Time series graph
In-Reply-To: <200810152021.6780@centrum.cz>
References: <200810152017.29054@centrum.cz> <200810152021.7089@centrum.cz>
	<200810152021.6780@centrum.cz>
Message-ID: <200810151604.20704.pgmdevlist@gmail.com>

Cenek,
Try the scikits.timeseries package 
(http://pytseries.sourceforge.net/installing.html).
One of its module (sciktis.timeseries.lib.plotlib) implements plotting 
capacities on top of matplotlib, and supports an update of ticks depending on 
the zoom level.


On Wednesday 15 October 2008 14:21:19 302302 wrote:
> Hi,
>
> I need to plot a time series (the data in the database has around 10000
> rows) and I need dynamically zoom and move with them. Something like
> http://finance.google.com/finance?q=INDEXDJX:.DJI%20INDEXNASDAQ:.IXIC%20IND
>EXSP:.INX Is there any efficient way how to do it with scipy or matplotlib?
>
> Thanks Cenek
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user




From strawman at astraw.com  Wed Oct 15 16:49:28 2008
From: strawman at astraw.com (Andrew Straw)
Date: Wed, 15 Oct 2008 13:49:28 -0700
Subject: [SciPy-user] Matlab can interfere with Python?
In-Reply-To: <6c476c8a0810151045wddcef6fgab1b98d9b4f8578a@mail.gmail.com>
References: <6c476c8a0810151045wddcef6fgab1b98d9b4f8578a@mail.gmail.com>
Message-ID: <48F65758.4000302@astraw.com>

Just to be clear, it seems like you have experienced problems with 2 things:

1) While doing the link step to create a .dll (well, .pyd file) when
running MinGW's GCC  on a .c file, the linking fails due to an error.
(The .c file being autogenerated by Cython when compiling a .pyx file.)
If this is correct, what do you mean by "prevent Python from linking" --
Python is not executing in this case. What is the exact gcc error
message? Does gcc not find the python25.dll? Does gcc fail to link other
files that have nothing to do with python (i.e. they don't use symbols
from python25.dll)?

2) "severly inhibit execution of code" -- are the symptoms any different
than the CPU being busy or the RAM being full? Have you checked those
things?

Still just trying to understand the nature of the problem,
Andrew

Wes McKinney wrote:
> I have experienced this bizarre problem on two my coworker's machine
> who run Matlab (7.6.0 rev 2008a) and I have been teaching to use
> Python/Numpy/Scipy. Occasionally Matlab will somehow prevent Python
> from linking with various DLLs or severely inhibit the execution of
> code. I am not really sure how this could be (other than the Borg at
> the MathWorks not wanting anyone to use open source software), but has
> anyone else experienced this? As soon as I killed Matlab the problem
> disappeared.
>
> Here is the linkage output from a Cython dll that was killed by this:
>
> C:\MinGW\bin\gcc.exe -mno-cygwin -shared -s
> build\temp.win32-2.5\Release\tseries.o
> build\temp.win32-2.5\Release\tseries.def -LC:\Python25\libs -LC:\P
> ython25\PCBuild -lpython25 -lmsvcr71 -o tseries.pyd
>
> Maybe it's locking up the windows C runtime msvcr71?
>
> Any help here would be appreciated.
>
> Thanks,
> Wes
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From wesmckinn at gmail.com  Wed Oct 15 18:21:27 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Wed, 15 Oct 2008 18:21:27 -0400
Subject: [SciPy-user] Matlab can interfere with Python?
In-Reply-To: <48F65758.4000302@astraw.com>
References: <6c476c8a0810151045wddcef6fgab1b98d9b4f8578a@mail.gmail.com>
	<48F65758.4000302@astraw.com>
Message-ID: <6c476c8a0810151521p3bca0c04m6ec32a45bb561110@mail.gmail.com>

There's nothing wrong with the DLL build =) I was experiencing problems with
unit-tested production code being ground to a halt when MATLAB is running in
the background. In other words, I would execute a simple line of code in the
Python shell and it would hang or take a very long time to execute. The
instant MATLAB was closed (matlab not executing any code, just idling),
Python stopped hanging and code would run as usual. (Seriously!)

I listed the gcc linkage since I thought Matlab might be interfering somehow
with the use of the Microsoft C runtime, but that doesn't make that much
sense either because Matlab has it's own C runtime DLL.

Thanks,
Wes

On Wed, Oct 15, 2008 at 4:49 PM, Andrew Straw <strawman at astraw.com> wrote:

> Just to be clear, it seems like you have experienced problems with 2
> things:
>
> 1) While doing the link step to create a .dll (well, .pyd file) when
> running MinGW's GCC  on a .c file, the linking fails due to an error.
> (The .c file being autogenerated by Cython when compiling a .pyx file.)
> If this is correct, what do you mean by "prevent Python from linking" --
> Python is not executing in this case. What is the exact gcc error
> message? Does gcc not find the python25.dll? Does gcc fail to link other
> files that have nothing to do with python (i.e. they don't use symbols
> from python25.dll)?
>
> 2) "severly inhibit execution of code" -- are the symptoms any different
> than the CPU being busy or the RAM being full? Have you checked those
> things?
>
> Still just trying to understand the nature of the problem,
> Andrew
>
> Wes McKinney wrote:
> > I have experienced this bizarre problem on two my coworker's machine
> > who run Matlab (7.6.0 rev 2008a) and I have been teaching to use
> > Python/Numpy/Scipy. Occasionally Matlab will somehow prevent Python
> > from linking with various DLLs or severely inhibit the execution of
> > code. I am not really sure how this could be (other than the Borg at
> > the MathWorks not wanting anyone to use open source software), but has
> > anyone else experienced this? As soon as I killed Matlab the problem
> > disappeared.
> >
> > Here is the linkage output from a Cython dll that was killed by this:
> >
> > C:\MinGW\bin\gcc.exe -mno-cygwin -shared -s
> > build\temp.win32-2.5\Release\tseries.o
> > build\temp.win32-2.5\Release\tseries.def -LC:\Python25\libs -LC:\P
> > ython25\PCBuild -lpython25 -lmsvcr71 -o tseries.pyd
> >
> > Maybe it's locking up the windows C runtime msvcr71?
> >
> > Any help here would be appreciated.
> >
> > Thanks,
> > Wes
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > SciPy-user mailing list
> > SciPy-user at scipy.org
> > http://projects.scipy.org/mailman/listinfo/scipy-user
> >
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081015/ba80e879/attachment.html>

From elcorto at gmx.net  Wed Oct 15 18:37:31 2008
From: elcorto at gmx.net (Steve Schmerler)
Date: Thu, 16 Oct 2008 00:37:31 +0200
Subject: [SciPy-user] npfile deprecated?
In-Reply-To: <3d375d730810151207i6efe0f6bwed94e65eeeae79ec@mail.gmail.com>
References: <20081015151203.GA7878@ramrod.starsheriffs.de>
	<3d375d730810151207i6efe0f6bwed94e65eeeae79ec@mail.gmail.com>
Message-ID: <20081015223731.GA8661@ramrod.starsheriffs.de>

On Oct 15 14:07 -0500, Robert Kern  wrote:
> On Wed, Oct 15, 2008 at 10:12, Steve Schmerler <elcorto at gmx.net> wrote:

[...]
> >
> > However, I found that in
> > the current svn HEAD, the scipy.io.npfile docstring says that it is
> > deprecated and I should use ndarray.tofile() and numpy.fromfile()
> > instead. But with the latter, I can't control byteorder, row- or
> > column-major order etc. That's a bit puzzling.
> >
> > I admitt that I haven't followed the development of the new numpy file
> > format very closely (since npfile worked nicely so far).  So, is the
> > .npy/.npz way the current standard to store platform-independent binary
> > arrays with all metainfo?
> 
> Yup.
> 

Ah OK, thanks. 

Regarding that docstrings, what about removing the lines

"You can achieve the same effect as using npfile, using ndarray.tofile
and numpy.fromfile."

from the np.deprecate_with_doc() text? The tofile/fromfile route doesn't
provide the same functionallity. It's confusing, since their docstrings
even discourage their usage for cross-platform case.

best,
steve


From robert.kern at gmail.com  Wed Oct 15 18:49:43 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 15 Oct 2008 17:49:43 -0500
Subject: [SciPy-user] npfile deprecated?
In-Reply-To: <20081015223731.GA8661@ramrod.starsheriffs.de>
References: <20081015151203.GA7878@ramrod.starsheriffs.de>
	<3d375d730810151207i6efe0f6bwed94e65eeeae79ec@mail.gmail.com>
	<20081015223731.GA8661@ramrod.starsheriffs.de>
Message-ID: <3d375d730810151549q525db796g848f29be0bd614f6@mail.gmail.com>

On Wed, Oct 15, 2008 at 17:37, Steve Schmerler <elcorto at gmx.net> wrote:
> On Oct 15 14:07 -0500, Robert Kern  wrote:
>> On Wed, Oct 15, 2008 at 10:12, Steve Schmerler <elcorto at gmx.net> wrote:
>
> [...]
>> >
>> > However, I found that in
>> > the current svn HEAD, the scipy.io.npfile docstring says that it is
>> > deprecated and I should use ndarray.tofile() and numpy.fromfile()
>> > instead. But with the latter, I can't control byteorder, row- or
>> > column-major order etc. That's a bit puzzling.
>> >
>> > I admitt that I haven't followed the development of the new numpy file
>> > format very closely (since npfile worked nicely so far).  So, is the
>> > .npy/.npz way the current standard to store platform-independent binary
>> > arrays with all metainfo?
>>
>> Yup.
>>
>
> Ah OK, thanks.
>
> Regarding that docstrings, what about removing the lines
>
> "You can achieve the same effect as using npfile, using ndarray.tofile
> and numpy.fromfile."
>
> from the np.deprecate_with_doc() text? The tofile/fromfile route doesn't
> provide the same functionallity. It's confusing, since their docstrings
> even discourage their usage for cross-platform case.

What functionality do you think that npfile provided which
tofile()/fromfile() doesn't? npfile never wrote out header
information. The files were always just the raw platform-specific
binary bytes. It's API is somewhat more convenient (you can specify
the shape in the method call rather than modifying it later), but
that's about it.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From strawman at astraw.com  Wed Oct 15 23:22:15 2008
From: strawman at astraw.com (Andrew Straw)
Date: Wed, 15 Oct 2008 20:22:15 -0700
Subject: [SciPy-user] Matlab can interfere with Python?
In-Reply-To: <6c476c8a0810151521p3bca0c04m6ec32a45bb561110@mail.gmail.com>
References: <6c476c8a0810151045wddcef6fgab1b98d9b4f8578a@mail.gmail.com>	<48F65758.4000302@astraw.com>
	<6c476c8a0810151521p3bca0c04m6ec32a45bb561110@mail.gmail.com>
Message-ID: <48F6B367.4080703@astraw.com>

Hmm, the only thing I can think of is that MATLAB has hold of some
resources and therefore the OS won't let Python use them. Is there
anything like lsof or strace one can use on Windows? Even putting print
statements all over your Python process and seeing what statement is
blocking execution would be a start. I'm afraid my Windows knowledge is
pretty minimal, so I can't be of much help here.

-Andrew

Wes McKinney wrote:
> There's nothing wrong with the DLL build =) I was experiencing
> problems with unit-tested production code being ground to a halt when
> MATLAB is running in the background. In other words, I would execute a
> simple line of code in the Python shell and it would hang or take a
> very long time to execute. The instant MATLAB was closed (matlab not
> executing any code, just idling), Python stopped hanging and code
> would run as usual. (Seriously!)
>
> I listed the gcc linkage since I thought Matlab might be interfering
> somehow with the use of the Microsoft C runtime, but that doesn't make
> that much sense either because Matlab has it's own C runtime DLL.
>
> Thanks,
> Wes
>
> On Wed, Oct 15, 2008 at 4:49 PM, Andrew Straw <strawman at astraw.com
> <mailto:strawman at astraw.com>> wrote:
>
>     Just to be clear, it seems like you have experienced problems with
>     2 things:
>
>     1) While doing the link step to create a .dll (well, .pyd file) when
>     running MinGW's GCC  on a .c file, the linking fails due to an error.
>     (The .c file being autogenerated by Cython when compiling a .pyx
>     file.)
>     If this is correct, what do you mean by "prevent Python from
>     linking" --
>     Python is not executing in this case. What is the exact gcc error
>     message? Does gcc not find the python25.dll? Does gcc fail to link
>     other
>     files that have nothing to do with python (i.e. they don't use symbols
>     from python25.dll)?
>
>     2) "severly inhibit execution of code" -- are the symptoms any
>     different
>     than the CPU being busy or the RAM being full? Have you checked those
>     things?
>
>     Still just trying to understand the nature of the problem,
>     Andrew
>
>     Wes McKinney wrote:
>     > I have experienced this bizarre problem on two my coworker's machine
>     > who run Matlab (7.6.0 rev 2008a) and I have been teaching to use
>     > Python/Numpy/Scipy. Occasionally Matlab will somehow prevent Python
>     > from linking with various DLLs or severely inhibit the execution of
>     > code. I am not really sure how this could be (other than the Borg at
>     > the MathWorks not wanting anyone to use open source software),
>     but has
>     > anyone else experienced this? As soon as I killed Matlab the problem
>     > disappeared.
>     >
>     > Here is the linkage output from a Cython dll that was killed by
>     this:
>     >
>     > C:\MinGW\bin\gcc.exe -mno-cygwin -shared -s
>     > build\temp.win32-2.5\Release\tseries.o
>     > build\temp.win32-2.5\Release\tseries.def -LC:\Python25\libs -LC:\P
>     > ython25\PCBuild -lpython25 -lmsvcr71 -o tseries.pyd
>     >
>     > Maybe it's locking up the windows C runtime msvcr71?
>     >
>     > Any help here would be appreciated.
>     >
>     > Thanks,
>     > Wes
>     >
>     ------------------------------------------------------------------------
>     >
>     > _______________________________________________
>     > SciPy-user mailing list
>     > SciPy-user at scipy.org <mailto:SciPy-user at scipy.org>
>     > http://projects.scipy.org/mailman/listinfo/scipy-user
>     >
>
>     _______________________________________________
>     SciPy-user mailing list
>     SciPy-user at scipy.org <mailto:SciPy-user at scipy.org>
>     http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From sahar at cmt.co.il  Thu Oct 16 02:53:08 2008
From: sahar at cmt.co.il (Sahar Vilan)
Date: Thu, 16 Oct 2008 08:53:08 +0200
Subject: [SciPy-user] saving raw image
In-Reply-To: <68B3EEDF-FB62-40B5-944A-D85B4E28A0B5@yale.edu>
Message-ID: <MNEGKHAFHCAFDOCPJMAMOEGKCBAA.sahar@cmt.co.il>

Hi Zach,
Thanks for your help.

I used your advices and got an image I can view. However, its size doubles
and I have to open it as 32 bit while the original looks fine as 16 bit.
Here is the code I used:
# ************************************************
from scipy import *

# read image
f0 = file('Image0.raw', 'rb').read()
Raw = fromstring(f0, uint16).astype(int)
Im = Raw.reshape([1024, 1024], order='F')

# save image
Im_str = Im.tostring(order='F')
f1 = file( 'Image1.raw', 'wb').write(Im_str)

# ************************************************
Thanks again,
Sahar

-----Original Message-----
From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org]On
Behalf Of Zachary Pincus
Sent: Wed, October 15, 2008 4:21 PM
To: SciPy Users List
Subject: Re: [SciPy-user] saving raw image


Hi Sahar,

Please send a minimal example of loading and saving an array that
produces "garbage" output, if you could... (If you could send a small
input file that would be helpful too.)

Also, does the garbage you see in ImageJ have some structure -- does
it look streaky, like you can see rows of pixels that should be
together, but the rows don't line up right?

One possible problem is that the array you are saving has been
promoted to a different dtype (e.g. by participating in signed/float
arithmetic), so the pixels you save are no longer uint16s. Unless you
have an 'astype(uint16)' in your save code, this could be the issue,
since the loading code you showed does immediately convert the array
from uint16s. (This is why sending a complete example of the failure
is useful...)

Another possible problem has to do with the order the pixels are read/
written out. Typically, images on disk are stored as rows of pixels
next to one another -- this is "column major" or "fortran" order
(going from one memory location to the next typically increments the x-
value, except at row boundaries where the y-value is incremented, so
it is said that the x-value "varies the fastest"). Typically, numpy
arrays are created in row major or "C" order, where the y-value varies
the fastest. When loading and manipulating images, you need either to
make sure that the images are loaded in fortran-order, or reverse the
x/y coordinates and shape.

So, e.g. when loading a 200x300 image:


s = file(path, 'rb').read()
raw = fromstring(s, uint16).astype(int)
image = raw.reshape((200,300), order='F')

now, image[30,40] gives the same pixel as coordinate (30,40) in ImageJ.

If you did:
image = raw.reshape((300,200), order='C')
then image[30,40] would give the same pixel as coordinate (40,30) in
ImageJ.

Note that the 'C' order is default. So:
image = raw.reshape((200,300))
will give garbage, with the rows of pixels broken up along the wrong
boundaries, giving rise to the "streaky" images I mentioned.

Finally, the tostring() method also takes an order option, so for
saving images, you need:
raw = image.tostring(order='F')
(assuming that you reshaped the image as order 'F')

Zach Pincus



On Oct 15, 2008, at 4:49 AM, Sahar Vilan wrote:

> I used scipy for basic image processing of raw images, and I can't
> save
> these images in the same format:
> To open image I use:
> 	s = file(path, 'rb').read()
>    	raw = fromstring(s, uint16).astype(int)
>
> I try to save this matrix as raw image again but I get some garbage
> when I
> open it in some viewer (Image-J, for instance).
> Can anyone help me with this?
>
> Thanks,
> Sahar

_______________________________________________
SciPy-user mailing list
SciPy-user at scipy.org
http://projects.scipy.org/mailman/listinfo/scipy-user



****************************************************************************
*************************** This e-mail message may contain confidential,and
privileged information or data that constitute proprietary information of
CMT Medical Ltd. Any review or distribution by others is strictly
prohibited. If you are not the intended recipient you are hereby notified
that any use of this information or data by any other person is absolutely
prohibited. If you are not the intended recipient, please delete all copies.
Thank You. http://www.cmt.co.il
****************************************************************************
****************************







****************************************************************************
********
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals & computer
viruses.
****************************************************************************
********




No virus found in this incoming message.
Checked by AVG - http://www.avg.com
Version: 8.0.173 / Virus Database: 270.8.0/1725 - Release Date: 14/10/2008
21:25




*******************************************************************************************************
This e-mail message may contain confidential,and  privileged information or data that constitute proprietary information of CMT Medical Ltd. Any review or distribution by others is strictly prohibited. If you are not the intended recipient you are hereby notified that any use of this information or data by any other person is absolutely prohibited. If you are not the intended recipient, please delete all copies. Thank You. http://www.cmt.co.il
********************************************************************************************************





 
 
************************************************************************************
This footnote confirms that this email message has been scanned by
PineApp Mail-SeCure for the presence of malicious code, vandals & computer viruses.
************************************************************************************





From wbaxter at gmail.com  Thu Oct 16 03:37:12 2008
From: wbaxter at gmail.com (Bill Baxter)
Date: Thu, 16 Oct 2008 16:37:12 +0900
Subject: [SciPy-user] CHOLMOD via scipy
Message-ID: <e86a5fd00810160037x2353c3f6j8a556267551110bb@mail.gmail.com>

Is CHOLMOD accessible through scipy somehow?
Or if not, is it available to python via some other easy-to-install package?

--bb


From cournape at gmail.com  Thu Oct 16 03:42:41 2008
From: cournape at gmail.com (David Cournapeau)
Date: Thu, 16 Oct 2008 16:42:41 +0900
Subject: [SciPy-user] Matlab can interfere with Python?
In-Reply-To: <6c476c8a0810151521p3bca0c04m6ec32a45bb561110@mail.gmail.com>
References: <6c476c8a0810151045wddcef6fgab1b98d9b4f8578a@mail.gmail.com>
	<48F65758.4000302@astraw.com>
	<6c476c8a0810151521p3bca0c04m6ec32a45bb561110@mail.gmail.com>
Message-ID: <5b8d13220810160042s19a5acd1yaf4ea034edc2516b@mail.gmail.com>

On Thu, Oct 16, 2008 at 7:21 AM, Wes McKinney <wesmckinn at gmail.com> wrote:
> There's nothing wrong with the DLL build =) I was experiencing problems with
> unit-tested production code being ground to a halt when MATLAB is running in
> the background. In other words, I would execute a simple line of code in the
> Python shell and it would hang or take a very long time to execute. The
> instant MATLAB was closed (matlab not executing any code, just idling),
> Python stopped hanging and code would run as usual. (Seriously!)
>
> I listed the gcc linkage since I thought Matlab might be interfering somehow
> with the use of the Microsoft C runtime, but that doesn't make that much
> sense either because Matlab has it's own C runtime DLL.

I sincerely doubt matlab has its own C runtime on windows. But even if
it is shared, if two processes using the same dll could interfere each
other: it would be an extremely serious bug in the core OS. Highly
unlikely.

Does it happen only with python ? Can you reproduce the hanging with a
small script ? The easiest way would be to be able to run the python
interpreter in the VS debugger, to see where it hangs.

cheers,

David


From schut at sarvision.nl  Thu Oct 16 04:39:26 2008
From: schut at sarvision.nl (Vincent Schut)
Date: Thu, 16 Oct 2008 10:39:26 +0200
Subject: [SciPy-user] saving raw image
In-Reply-To: <MNEGKHAFHCAFDOCPJMAMOEGKCBAA.sahar@cmt.co.il>
References: <68B3EEDF-FB62-40B5-944A-D85B4E28A0B5@yale.edu>
	<MNEGKHAFHCAFDOCPJMAMOEGKCBAA.sahar@cmt.co.il>
Message-ID: <gd6uju$cr6$1@ger.gmane.org>

Numpy int's are int32 by default, so in your '.astype(int)' you're 
casting to int32. If you want a different type, you need to specify the 
number of bits. As a side note, you're reading as *unsigned* integer 
(uint16) and then converting to *signed* int32. Try "Raw = 
fromstring(f0, uint16).astype(uint16)" to keep the original (unsigned 
int16) datatype.

Vincent.

Sahar Vilan wrote:
> Hi Zach,
> Thanks for your help.
> 
> I used your advices and got an image I can view. However, its size doubles
> and I have to open it as 32 bit while the original looks fine as 16 bit.
> Here is the code I used:
> # ************************************************
> from scipy import *
> 
> # read image
> f0 = file('Image0.raw', 'rb').read()
> Raw = fromstring(f0, uint16).astype(int)
> Im = Raw.reshape([1024, 1024], order='F')
> 
> # save image
> Im_str = Im.tostring(order='F')
> f1 = file( 'Image1.raw', 'wb').write(Im_str)
> 
> # ************************************************
> Thanks again,
> Sahar
> 
> -----Original Message-----
> From: scipy-user-bounces at scipy.org [mailto:scipy-user-bounces at scipy.org]On
> Behalf Of Zachary Pincus
> Sent: Wed, October 15, 2008 4:21 PM
> To: SciPy Users List
> Subject: Re: [SciPy-user] saving raw image
> 
> 
> Hi Sahar,
> 
> Please send a minimal example of loading and saving an array that
> produces "garbage" output, if you could... (If you could send a small
> input file that would be helpful too.)
> 
> Also, does the garbage you see in ImageJ have some structure -- does
> it look streaky, like you can see rows of pixels that should be
> together, but the rows don't line up right?
> 
> One possible problem is that the array you are saving has been
> promoted to a different dtype (e.g. by participating in signed/float
> arithmetic), so the pixels you save are no longer uint16s. Unless you
> have an 'astype(uint16)' in your save code, this could be the issue,
> since the loading code you showed does immediately convert the array
> from uint16s. (This is why sending a complete example of the failure
> is useful...)
> 
> Another possible problem has to do with the order the pixels are read/
> written out. Typically, images on disk are stored as rows of pixels
> next to one another -- this is "column major" or "fortran" order
> (going from one memory location to the next typically increments the x-
> value, except at row boundaries where the y-value is incremented, so
> it is said that the x-value "varies the fastest"). Typically, numpy
> arrays are created in row major or "C" order, where the y-value varies
> the fastest. When loading and manipulating images, you need either to
> make sure that the images are loaded in fortran-order, or reverse the
> x/y coordinates and shape.
> 
> So, e.g. when loading a 200x300 image:
> 
> 
> s = file(path, 'rb').read()
> raw = fromstring(s, uint16).astype(int)
> image = raw.reshape((200,300), order='F')
> 
> now, image[30,40] gives the same pixel as coordinate (30,40) in ImageJ.
> 
> If you did:
> image = raw.reshape((300,200), order='C')
> then image[30,40] would give the same pixel as coordinate (40,30) in
> ImageJ.
> 
> Note that the 'C' order is default. So:
> image = raw.reshape((200,300))
> will give garbage, with the rows of pixels broken up along the wrong
> boundaries, giving rise to the "streaky" images I mentioned.
> 
> Finally, the tostring() method also takes an order option, so for
> saving images, you need:
> raw = image.tostring(order='F')
> (assuming that you reshaped the image as order 'F')
> 
> Zach Pincus
> 
> 
> 
> On Oct 15, 2008, at 4:49 AM, Sahar Vilan wrote:
> 
>> I used scipy for basic image processing of raw images, and I can't
>> save
>> these images in the same format:
>> To open image I use:
>> 	s = file(path, 'rb').read()
>>    	raw = fromstring(s, uint16).astype(int)
>>
>> I try to save this matrix as raw image again but I get some garbage
>> when I
>> open it in some viewer (Image-J, for instance).
>> Can anyone help me with this?
>>
>> Thanks,
>> Sahar
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
> 
> 
> 
> ****************************************************************************
> *************************** This e-mail message may contain confidential,and
> privileged information or data that constitute proprietary information of
> CMT Medical Ltd. Any review or distribution by others is strictly
> prohibited. If you are not the intended recipient you are hereby notified
> that any use of this information or data by any other person is absolutely
> prohibited. If you are not the intended recipient, please delete all copies.
> Thank You. http://www.cmt.co.il
> ****************************************************************************
> ****************************
> 
> 
> 
> 
> 
> 
> 
> ****************************************************************************
> ********
> This footnote confirms that this email message has been scanned by
> PineApp Mail-SeCure for the presence of malicious code, vandals & computer
> viruses.
> ****************************************************************************
> ********
> 
> 
> 
> 
> No virus found in this incoming message.
> Checked by AVG - http://www.avg.com
> Version: 8.0.173 / Virus Database: 270.8.0/1725 - Release Date: 14/10/2008
> 21:25
> 
> 
> 
> 
> *******************************************************************************************************
> This e-mail message may contain confidential,and  privileged information or data that constitute proprietary information of CMT Medical Ltd. Any review or distribution by others is strictly prohibited. If you are not the intended recipient you are hereby notified that any use of this information or data by any other person is absolutely prohibited. If you are not the intended recipient, please delete all copies. Thank You. http://www.cmt.co.il
> ********************************************************************************************************
> 
> 
> 
> 
> 
>  
>  
> ************************************************************************************
> This footnote confirms that this email message has been scanned by
> PineApp Mail-SeCure for the presence of malicious code, vandals & computer viruses.
> ************************************************************************************



From tmspriy at nus.edu.sg  Thu Oct 16 04:53:42 2008
From: tmspriy at nus.edu.sg (priya_tmsi)
Date: Thu, 16 Oct 2008 01:53:42 -0700 (PDT)
Subject: [SciPy-user]  signal.medfilt
Message-ID: <20009033.post@talk.nabble.com>


I am trying to write a code for median filter using scipy
the following is my code

from numpy import *
import ImageFilter
from scipy import *
import scipy.signal as signal
img = file('girl.png','rb').read() 
imgDest=signal.medfilt(img)
f1 = file( 'pygirl.png').write(imgDest)

I get the following error on running this code

Traceback (most recent call last):
  File "medfilt.py", line 6, in ?
    imgDest=signal.medfilt(img)
  File "/usr/lib/python2.4/site-packages/scipy/signal/signaltools.py", line
232, in medfilt
    return sigtools._order_filterND(volume,domain,order)
ValueError: data type must provide an itemsize


Can any one help please...:,(:,(

-- 
View this message in context: http://www.nabble.com/signal.medfilt-tp20009033p20009033.html
Sent from the Scipy-User mailing list archive at Nabble.com.



From nwagner at iam.uni-stuttgart.de  Thu Oct 16 06:08:15 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 16 Oct 2008 12:08:15 +0200
Subject: [SciPy-user] CHOLMOD via scipy
In-Reply-To: <e86a5fd00810160037x2353c3f6j8a556267551110bb@mail.gmail.com>
References: <e86a5fd00810160037x2353c3f6j8a556267551110bb@mail.gmail.com>
Message-ID: <web-112919396@uni-stuttgart.de>

On Thu, 16 Oct 2008 16:37:12 +0900
  "Bill Baxter" <wbaxter at gmail.com> wrote:
> Is CHOLMOD accessible through scipy somehow?
> Or if not, is it available to python via some other 
>easy-to-install package?
> 
> --bb
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

Hi Bill,

It is on the "Wishlist" ...
http://projects.scipy.org/scipy/scipy/ticket/261

You may want to use cvxopt in the meantime

See
http://abel.ee.ucla.edu/cvxopt/documentation/users-guide/node41.html
  
Cheers,
            Nils


From zachary.pincus at yale.edu  Thu Oct 16 08:14:54 2008
From: zachary.pincus at yale.edu (Zachary Pincus)
Date: Thu, 16 Oct 2008 08:14:54 -0400
Subject: [SciPy-user] saving raw image
In-Reply-To: <gd6uju$cr6$1@ger.gmane.org>
References: <68B3EEDF-FB62-40B5-944A-D85B4E28A0B5@yale.edu>
	<MNEGKHAFHCAFDOCPJMAMOEGKCBAA.sahar@cmt.co.il>
	<gd6uju$cr6$1@ger.gmane.org>
Message-ID: <A6A2A111-688B-4014-9AB3-FC6292007683@yale.edu>

> Raw = fromstring(f0, uint16).astype(uint16)

Note that this is a tad unnecessary -- fromstring produces arrays with  
the requested dtype:

In : numpy.fromstring('abcd', dtype=numpy.uint16)
Out: array([25185, 25699], dtype=uint16)

However, it is easy to accidentally promote the output of math  
operations to different dtypes in math operations with signed int  
types or floats.

You best bet is to make sure that you convert the image back to 16-bit  
unsigned ints before saving:

# save image
Im_str = Im.astype(numpy.uint16).tostring(order='F')
f1 = file( 'Image1.raw', 'wb').write(Im_str)

Note that this conversion works like C casting, so if you've got  
negative values you will have wraparound.

Zach


From elcorto at gmx.net  Thu Oct 16 08:15:07 2008
From: elcorto at gmx.net (Steve Schmerler)
Date: Thu, 16 Oct 2008 14:15:07 +0200
Subject: [SciPy-user] npfile deprecated?
In-Reply-To: <3d375d730810151549q525db796g848f29be0bd614f6@mail.gmail.com>
References: <20081015151203.GA7878@ramrod.starsheriffs.de>
	<3d375d730810151207i6efe0f6bwed94e65eeeae79ec@mail.gmail.com>
	<20081015223731.GA8661@ramrod.starsheriffs.de>
	<3d375d730810151549q525db796g848f29be0bd614f6@mail.gmail.com>
Message-ID: <20081016121507.GA13628@ramrod.starsheriffs.de>

On Oct 15 17:49 -0500, Robert Kern  wrote:
> What functionality do you think that npfile provided which
> tofile()/fromfile() doesn't? npfile never wrote out header
> information. The files were always just the raw platform-specific
> binary bytes. It's API is somewhat more convenient (you can specify
> the shape in the method call rather than modifying it later), but
> that's about it.
> 

Well ok, it's mostly only the API rather than functionallity.

I suppose that with tofile/fromfile, I could use approximately this
instead:
    
    >>> sys.byteorder
    'little' 
    >>> sh = (3,5)
    >>> ord = 'F'
    >>> a = asarray(np.ramdom.random(sh), order=ord)
    >>> dt = a.dtype
    >>> a.tofile('test.dat')

and on another machine
    
    >>> a = fromfile('test.dat', dtype=dt).reshape(sh)
    >>> a = asarray(a, order='F')
    >>> if sys.byteorder == 'big': a = a.byteswap()

Here, and in the npfile case, one would use wrappers writing and reading
metainfos (dtype, order, byteorder and shape) to and from an extra file
(at least, I did this).

The only functionallity difference is that files on disk would be always
in C-order. 

Anyway, god bless numpy.save() :-)

best,
steve


From zachary.pincus at yale.edu  Thu Oct 16 08:22:14 2008
From: zachary.pincus at yale.edu (Zachary Pincus)
Date: Thu, 16 Oct 2008 08:22:14 -0400
Subject: [SciPy-user] signal.medfilt
In-Reply-To: <20009033.post@talk.nabble.com>
References: <20009033.post@talk.nabble.com>
Message-ID: <DB2EDD8A-3024-4D77-A447-2C08363C737E@yale.edu>

Hi Priya,

I think the problem may be that you aren't properly reading in the  
images... unless you've got some other code that is overriding the  
builtin python 'file' function, then the following simply reads in the  
bits from a png file into a string, without decoding/decompressing/etc  
the file into a numpy array:

> img = file('girl.png','rb').read()

As such, it's no surprise that a function which works on numpy arrays  
is choking when fed a string.

You'll need to get some sort of image IO library to read PNGs. My  
suggestion would be to look into PIL, the Python Imaging Library,  
which can read images and be used to produce numpy arrays from them.

Once you are able to read in images, I also suggest looking at  
scipy.ndimage, which provides useful image manipulation tools.

Zach


On Oct 16, 2008, at 4:53 AM, priya_tmsi wrote:

>
> I am trying to write a code for median filter using scipy
> the following is my code
>
> from numpy import *
> import ImageFilter
> from scipy import *
> import scipy.signal as signal
> img = file('girl.png','rb').read()
> imgDest=signal.medfilt(img)
> f1 = file( 'pygirl.png').write(imgDest)
>
> I get the following error on running this code
>
> Traceback (most recent call last):
>  File "medfilt.py", line 6, in ?
>    imgDest=signal.medfilt(img)
>  File "/usr/lib/python2.4/site-packages/scipy/signal/ 
> signaltools.py", line
> 232, in medfilt
>    return sigtools._order_filterND(volume,domain,order)
> ValueError: data type must provide an itemsize
>
>
> Can any one help please...:,(:,(
>
> -- 
> View this message in context: http://www.nabble.com/signal.medfilt-tp20009033p20009033.html
> Sent from the Scipy-User mailing list archive at Nabble.com.
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From cohen at slac.stanford.edu  Thu Oct 16 12:35:42 2008
From: cohen at slac.stanford.edu (Johann Cohen-Tanugi)
Date: Thu, 16 Oct 2008 18:35:42 +0200
Subject: [SciPy-user] scipy 0.6.0 build failing
In-Reply-To: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>
References: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>
Message-ID: <48F76D5E.9050708@slac.stanford.edu>

I am not a specialist, but the first thing I would do is delete the 
local build directory, and then retry using the --fcompiler=gnu95 
argument to python setup.py build, so that gfortran is picked up instead 
of g77.
HTH,
Johann

chris wrote:
> I'm trying to build scipy 0.6.0 on RHEL 3, and am getting the following failure:
>
>
> g77:f77: scipy/fftpack/dfftpack/zfftf1.f
> /tmp/cceGs6VT.s: Assembler messages:
> /tmp/cceGs6VT.s:598: Error: suffix or operands invalid for `movd'
> /tmp/cceGs6VT.s:2994: Error: suffix or operands invalid for `movd'
> /tmp/cceGs6VT.s: Assembler messages:
> /tmp/cceGs6VT.s:598: Error: suffix or operands invalid for `movd'
> /tmp/cceGs6VT.s:2994: Error: suffix or operands invalid for `movd'
> error: Command "/usr/bin/g77 -g -Wall -fno-second-underscore -fPIC -O2
> -funroll-loops -march=i686 -mmmx -msse2 -msse -fomit-frame-pointer
> -malign-double -c -c scipy/fftpack/dfftpack/zfftf1.f -o
> build/temp.linux-i686-2.5/scipy/fftpack/dfftpack/zfftf1.o" failed with
> exit status 1
>
>
> g77 version 3.2.3
> gcc version 3.2.3
> as version 2.14.90.0.4
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   


From cournape at gmail.com  Fri Oct 17 01:16:33 2008
From: cournape at gmail.com (David Cournapeau)
Date: Fri, 17 Oct 2008 14:16:33 +0900
Subject: [SciPy-user] scipy 0.6.0 build failing
In-Reply-To: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>
References: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>
Message-ID: <5b8d13220810162216x96b9c8ek3648fed2724fe999@mail.gmail.com>

On Tue, Oct 14, 2008 at 3:09 AM, chris <cdcasey at gmail.com> wrote:
> I'm trying to build scipy 0.6.0 on RHEL 3, and am getting the following failure:
>
>
> g77:f77: scipy/fftpack/dfftpack/zfftf1.f
> /tmp/cceGs6VT.s: Assembler messages:
> /tmp/cceGs6VT.s:598: Error: suffix or operands invalid for `movd'

Your version of g77 has a bug, and generate invalid machine
instructions when the -msse option is used. You should update your g77
if possible,

David


From travis at enthought.com  Fri Oct 17 12:35:03 2008
From: travis at enthought.com (Travis Vaught)
Date: Fri, 17 Oct 2008 11:35:03 -0500
Subject: [SciPy-user] Time series graph
In-Reply-To: <200810151604.20704.pgmdevlist@gmail.com>
References: <200810152017.29054@centrum.cz> <200810152021.7089@centrum.cz>
	<200810152021.6780@centrum.cz>
	<200810151604.20704.pgmdevlist@gmail.com>
Message-ID: <AB3DE53C-4750-4179-8C0A-B11190DF8631@enthought.com>


On Oct 15, 2008, at 3:04 PM, Pierre GM wrote:

> Cenek,
> Try the scikits.timeseries package
> (http://pytseries.sourceforge.net/installing.html).
> One of its module (sciktis.timeseries.lib.plotlib) implements plotting
> capacities on top of matplotlib, and supports an update of ticks  
> depending on
> the zoom level.
>
>
> On Wednesday 15 October 2008 14:21:19 302302 wrote:
>> Hi,
>>
>> I need to plot a time series (the data in the database has around  
>> 10000
>> rows) and I need dynamically zoom and move with them. Something like
>> http://finance.google.com/finance?q=INDEXDJX:.DJI%20INDEXNASDAQ:.IXIC%20IND
>> EXSP:.INX Is there any efficient way how to do it with scipy or  
>> matplotlib?
>>
>> Thanks Cenek
>>
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>

Hey Cenek,

There's a chaco example of this here:

https://svn.enthought.com/enthought/browser/Chaco/trunk/examples/financial_plot_dates.py

It gives you zoomability and very nice tick handling of dates:

-------------- next part --------------
A non-text attachment was scrubbed...
Name: pastedGraphic.png
Type: image/png
Size: 30249 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081017/1583d1dc/attachment.png>
-------------- next part --------------



Zooming in quite a bit gives tick labels at the hour or minute scales  
(with reasonable scales in between):

-------------- next part --------------
A non-text attachment was scrubbed...
Name: pastedGraphic.png
Type: image/png
Size: 25954 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081017/1583d1dc/attachment-0001.png>
-------------- next part --------------



HTH,

Travis

From olli.sipila at helsinki.fi  Sat Oct 18 08:06:41 2008
From: olli.sipila at helsinki.fi (Olli =?utf-8?b?U2lwaWzDpA==?=)
Date: Sat, 18 Oct 2008 15:06:41 +0300
Subject: [SciPy-user] scipy 0.6 build error
Message-ID: <20081018150641.17457hwrt91wafkh@webmail.helsinki.fi>

I'm having a problem installing scipy on OS X Tiger (10.4.11). To be  
more specific, I'm trying to install the package manually according to  
the instructions on the scipy.org webpage; the building process goes  
smoothly until I get the following error message:

compile options: '-DNO_ATLAS_INFO=3 -DUSE_VENDOR_BLAS=1  
-I/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/numpy/core/include -I/Library/Frameworks/Python.framework/Versions/2.6/include/python2.6  
-c'
extra options: '-faltivec'
gcc: scipy/linsolve/_superluobject.c
In file included from scipy/linsolve/_superluobject.h:8,
                  from scipy/linsolve/_superluobject.c:5:
scipy/linsolve/SuperLU/SRC/scomplex.h:60: error: conflicting types for  
'_Py_c_abs'
/Library/Frameworks/Python.framework/Versions/2.6/include/python2.6/complexobject.h:30: error: previous declaration of '_Py_c_abs' was  
here
In file included from scipy/linsolve/_superluobject.h:8,
                  from scipy/linsolve/_superluobject.c:5:
scipy/linsolve/SuperLU/SRC/scomplex.h:60: error: conflicting types for  
'_Py_c_abs'
/Library/Frameworks/Python.framework/Versions/2.6/include/python2.6/complexobject.h:30: error: previous declaration of '_Py_c_abs' was  
here
lipo: can't figure out the architecture type of: /var/tmp//ccFsb10y.out
In file included from scipy/linsolve/_superluobject.h:8,
                  from scipy/linsolve/_superluobject.c:5:
scipy/linsolve/SuperLU/SRC/scomplex.h:60: error: conflicting types for  
'_Py_c_abs'
/Library/Frameworks/Python.framework/Versions/2.6/include/python2.6/complexobject.h:30: error: previous declaration of '_Py_c_abs' was  
here
In file included from scipy/linsolve/_superluobject.h:8,
                  from scipy/linsolve/_superluobject.c:5:
scipy/linsolve/SuperLU/SRC/scomplex.h:60: error: conflicting types for  
'_Py_c_abs'
/Library/Frameworks/Python.framework/Versions/2.6/include/python2.6/complexobject.h:30: error: previous declaration of '_Py_c_abs' was  
here
lipo: can't figure out the architecture type of: /var/tmp//ccFsb10y.out
error: Command "gcc -arch ppc -arch i386 -isysroot  
/Developer/SDKs/MacOSX10.4u.sdk -fno-strict-aliasing -fno-common  
-dynamic -DNDEBUG -g -O3 -DNO_ATLAS_INFO=3 -DUSE_VENDOR_BLAS=1  
-I/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/numpy/core/include -I/Library/Frameworks/Python.framework/Versions/2.6/include/python2.6 -c scipy/linsolve/_superluobject.c -o build/temp.macosx-10.4-fat-2.6/scipy/linsolve/_superluobject.o -faltivec" failed with exit status  
1

If it is indeed the -faltivec option that is causing this error, is  
there any way to disable it? My configuration at the moment is
MAC OS X Tiger (10.4.11)
gcc version 4.0.1
gfortran version 4.2.3 (obtained from the link at scipy.org)

The version of scipy I'm trying to build is 0.6. If someone has any  
thoughts on how to fix this, it would be very much appreciated.

- Olli Sipil?



From contact at pythonxy.com  Sun Oct 19 09:35:31 2008
From: contact at pythonxy.com (Pierre Raybaut)
Date: Sun, 19 Oct 2008 15:35:31 +0200
Subject: [SciPy-user] Reinteract
Message-ID: <48FB37A3.2010304@pythonxy.com>

Hi all,

A few months ago, I've discovered Reinteract 
(http://www.reinteract.org), a promising project providing an 
interactive Python environment based on worksheets (it reminds me 
somehow of Mathematica's).
Now Reinteract has a lot more features, and even if the developer claims 
that it's still an early stage development release (0.4.0), it's clearly 
worth a try.
Look at this screencast, it's very impressive:
http://www.gnome.org/~otaylor/reinteract-demo.html

So, if you're interested in trying Reinteract on Windows, I invite you 
to download the Reinteract and PyGTK (Reinteract's backend) Python(x,y) 
plugins:
http://www.pythonxy.com/additional.php#third

Note that even if a Python(x,y) install is recommended (e.g. for a 
better help integration and update management), you may install a 
Python(x,y) plugin over a standard Python 2.5 install (i.e. if you have 
installed Python 2.5 with the official .msi installer).

Cheers,
Pierre


From pgmdevlist at gmail.com  Sun Oct 19 19:43:16 2008
From: pgmdevlist at gmail.com (Pierre GM)
Date: Sun, 19 Oct 2008 19:43:16 -0400
Subject: [SciPy-user] Record Array: How to add a column?
In-Reply-To: <88e473830810140955u6f5d830cueb24a3f4bc980458@mail.gmail.com>
References: <48F3E642.4040001@cornell.edu>
	<200810141228.42850.pgmdevlist@gmail.com>
	<88e473830810140955u6f5d830cueb24a3f4bc980458@mail.gmail.com>
Message-ID: <200810191943.16946.pgmdevlist@gmail.com>

John,
Could you give the SVN>=r5949  a try ? That fixes one of the problems you were 
describing in a previous email:
r2[0].date.year
The reason why it didn't work before was that a masked array was 
systematically returned in __getattribute__. Now, you'll get a masked array 
if the output has a shape (ie, not a single record), or if it's a single 
record with some mask fields.

Note that I'm not guaranteeing that you won't run into some problems again at 
one point or another...



From cournape at gmail.com  Sun Oct 19 22:00:08 2008
From: cournape at gmail.com (David Cournapeau)
Date: Mon, 20 Oct 2008 11:00:08 +0900
Subject: [SciPy-user] scipy 0.6 build error
In-Reply-To: <20081018150641.17457hwrt91wafkh@webmail.helsinki.fi>
References: <20081018150641.17457hwrt91wafkh@webmail.helsinki.fi>
Message-ID: <5b8d13220810191900i7d511b25vc6bb678ef821a97a@mail.gmail.com>

On Sat, Oct 18, 2008 at 9:06 PM, Olli Sipil? <olli.sipila at helsinki.fi> wrote:
>
> The version of scipy I'm trying to build is 0.6. If someone has any
> thoughts on how to fix this, it would be very much appreciated.

You should not use python 2.6 for the time being. Numpy and scipy
cannot not (yet) be considered as 2.6 compatible.

cheers,

David

From olli.sipila at helsinki.fi  Mon Oct 20 02:51:06 2008
From: olli.sipila at helsinki.fi (Olli =?utf-8?b?U2lwaWzDpA==?=)
Date: Mon, 20 Oct 2008 09:51:06 +0300
Subject: [SciPy-user] scipy 0.6 build error
In-Reply-To: <5b8d13220810191900i7d511b25vc6bb678ef821a97a@mail.gmail.com>
References: <20081018150641.17457hwrt91wafkh@webmail.helsinki.fi>
	<5b8d13220810191900i7d511b25vc6bb678ef821a97a@mail.gmail.com>
Message-ID: <20081020095106.58557anc9o1evwwq@webmail.helsinki.fi>


Hm ok, I'll give it a go. I recall 2.5 also giving me an error message  
(this I think is the reason why I wanted to try 2.6), but I'll post  
that error here as well if 2.5 doesn't work either.

- Olli

Quoting "David Cournapeau" <cournape at gmail.com>:

> On Sat, Oct 18, 2008 at 9:06 PM, Olli Sipil? <olli.sipila at helsinki.fi> wrote:
>>
>> The version of scipy I'm trying to build is 0.6. If someone has any
>> thoughts on how to fix this, it would be very much appreciated.
>
> You should not use python 2.6 for the time being. Numpy and scipy
> cannot not (yet) be considered as 2.6 compatible.
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>




From lionel at gamr7.com  Mon Oct 20 05:37:06 2008
From: lionel at gamr7.com (Lionel Barret De Nazaris)
Date: Mon, 20 Oct 2008 11:37:06 +0200
Subject: [SciPy-user] Arc-length reparametrization. (newbie question ?)
Message-ID: <48FC5142.508@gamr7.com>

Hello all,

I've just inherited from a bunch of not so-good code about cubic
splines, and as a relative scipy newbie, I was wondering what would be 
the right way to do it.

I need to do Arc-length reparametrization.( 
http://www.math.hmc.edu/~gu/math142/mellon/Differential_Geometry/Geometry_of_curves/Parametric_Curves_and_arc.html 
<http://www.math.hmc.edu/%7Egu/math142/mellon/Differential_Geometry/Geometry_of_curves/Parametric_Curves_and_arc.html> 
)

In the current code, I recognize the use of the newton's method to 
converge on a good estimate but this leave me non-plussed. The heavy use 
of integration to compute the curve length without keeping the 
intermediary results is really weird.

So how would you do it, using the best of what scipy has to offer ?

Note : I've looked as the splines in scipy , but this seems more focused 
on finding the spline that fits the samples, than on this kind of 
manipulation.
Here we have the spline and its control points.

this the code  (see http://pastebin.com/m7e276652 if indentation is not 
correct) :

<code>
def derivated_curvilign_abscissa(t, points):
   return 
Vector(vectorial_derivated_cubic_interpolator(t,points)).magnitude()

def curvilign_abscissa(t, points):
   """
    input :
      points = spline control points [init_point, init_virtual_point, 
end_virtual_point, end_point]
      t      = a float between 0 and 1
    ouput :
      the arc-length between f(t) and the start of the curve (aka 
curvilign_abscissa)
    """
   return Vector(points[0]).magnitude() + quad(lambda x 
:derivated_curvilign_abscissa( x, points), 0., t)[0]

def create_curvy_grid(points, samples):
   """
   input :
      points  = spline control points [init_point, init_virtual_point, 
end_virtual_point, end_point]
      samples = list of of equally spaced t (float) between 0 and 1
   output :
       a list of float t  (float) where moving from tn to tn+1 means an 
advancing on the curve of an equal length s.
   """
   curve_length = curvilign_abscissa(1., points) -curvilign_abscissa(0., 
points)
   # ==
   def function_to_solve(x, translation) :
       return - translation + curvilign_abscissa(x, points) 
-curvilign_abscissa(0., points)
   # ==
   return [newton(lambda x : function_to_solve(x, curve_length * 
step_grid), 1 - PRECISION) for step_grid in grid]
   # ==
</code>




From david at ar.media.kyoto-u.ac.jp  Mon Oct 20 07:48:55 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Mon, 20 Oct 2008 20:48:55 +0900
Subject: [SciPy-user] scipy 0.6 build error
In-Reply-To: <20081020095106.58557anc9o1evwwq@webmail.helsinki.fi>
References: <20081018150641.17457hwrt91wafkh@webmail.helsinki.fi>	<5b8d13220810191900i7d511b25vc6bb678ef821a97a@mail.gmail.com>
	<20081020095106.58557anc9o1evwwq@webmail.helsinki.fi>
Message-ID: <48FC7027.9010502@ar.media.kyoto-u.ac.jp>

Olli Sipil? wrote:
> Hm ok, I'll give it a go. I recall 2.5 also giving me an error message  
> (this I think is the reason why I wanted to try 2.6), but I'll post  
> that error here as well if 2.5 doesn't work either.
>   

Yes, please report any error on 2.5. Note that even if the error is not
specific to 2.6, you should not (yet) use numpy and/or scipy with 2.6.

cheers,

David


From josegomez at gmx.net  Mon Oct 20 10:05:40 2008
From: josegomez at gmx.net (Jose Luis Gomez Dans)
Date: Mon, 20 Oct 2008 16:05:40 +0200
Subject: [SciPy-user] Best way to test several values
Message-ID: <20081020140540.5360@gmx.net>

Hi,
I have a 2D array of numbers, and I want to use "where" to effectively mask
bits of the array out. Essentially, if any element in this 2D array of
numbers belongs to a given list (stored as a 1D array), the condition should
be e.g., True.

Up to now, I have used numpy.any ( data==QA ), where data is a single element
of my 2D array, and QA is my acceptance values list. However, I need to loop
through elements, and is taking a long time.

Any hints? numpy/scipy seems to have a plethora of clever functions to this
sort of thing efficiently!

Thanks
Jose
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser


From robert.kern at gmail.com  Mon Oct 20 10:15:06 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 20 Oct 2008 09:15:06 -0500
Subject: [SciPy-user] Best way to test several values
In-Reply-To: <20081020140540.5360@gmx.net>
References: <20081020140540.5360@gmx.net>
Message-ID: <3d375d730810200715y31cf878erf1a3a88302dc768d@mail.gmail.com>

On Mon, Oct 20, 2008 at 09:05, Jose Luis Gomez Dans <josegomez at gmx.net> wrote:
> Hi,
> I have a 2D array of numbers, and I want to use "where" to effectively mask
> bits of the array out. Essentially, if any element in this 2D array of
> numbers belongs to a given list (stored as a 1D array), the condition should
> be e.g., True.
>
> Up to now, I have used numpy.any ( data==QA ), where data is a single element
> of my 2D array, and QA is my acceptance values list. However, I need to loop
> through elements, and is taking a long time.
>
> Any hints? numpy/scipy seems to have a plethora of clever functions to this
> sort of thing efficiently!

In [12]: a = random.randint(0, 10, (10,10))

In [13]: b = reshape(a, a.shape + (1,))

In [14]: QA = array([2,4,6])

In [15]: QA2 = reshape(QA, (1,1,-1))

In [16]: (b == QA2).any(axis=-1)
Out[16]:
array([[ True,  True, False, False,  True, False, False, False, False,
         True],
       [ True, False, False, False,  True,  True,  True, False, False,
        False],
       [ True,  True, False, False,  True,  True, False,  True,  True,
        False],
       [ True, False, False, False, False, False,  True, False,  True,
        False],
       [False,  True, False, False,  True, False, False, False,  True,
         True],
       [False, False,  True, False, False, False, False,  True, False,
        False],
       [False, False,  True,  True,  True,  True, False, False, False,
        False],
       [ True, False,  True, False, False, False, False, False, False,
        False],
       [False, False, False, False, False, False, False, False, False,
         True],
       [ True, False, False, False, False, False, False, False,  True,
         True]], dtype=bool)

In [17]: a[Out[16]]
Out[17]:
array([6, 2, 6, 4, 6, 4, 6, 2, 2, 4, 2, 6, 2, 2, 6, 4, 2, 6, 4, 2, 6, 4, 6,
       2, 6, 2, 6, 4, 4, 2, 2, 4, 4])

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From aisaac at american.edu  Mon Oct 20 20:54:40 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Mon, 20 Oct 2008 20:54:40 -0400
Subject: [SciPy-user] where is models.py?
Message-ID: <48FD2850.5090308@american.edu>

models.py seems to have been moved?
Alan Isaac



From millman at berkeley.edu  Mon Oct 20 23:02:38 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Mon, 20 Oct 2008 20:02:38 -0700
Subject: [SciPy-user] where is models.py?
In-Reply-To: <48FD2850.5090308@american.edu>
References: <48FD2850.5090308@american.edu>
Message-ID: <c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>

On Mon, Oct 20, 2008 at 5:54 PM, Alan G Isaac <aisaac at american.edu> wrote:
> models.py seems to have been moved?

Assuming you are referring to the models package, I removed it in August:
http://projects.scipy.org/scipy/scipy/changeset/4639
(If not, sorry for the confusion.)

I sent an email to the list before hand explaining that
scipy.stats.models wasn't ready for release and proposed removing it
so that we could move forward on the 0.7 release.  I haven't looked at
the code since then, but as I recall there were several broken tests
and some open questions about what the API should be.  I spoke with
Jonathan Taylor (he wrote the package) and he said he wasn't going to
have time in the near future to work on it.

This is very unfortunate since the statistical models package provides
some really useful functionality and I would love to see it in the 0.7
release.  For the time being, it is living (and hopefully developing)
in the nipy project:  https://launchpad.net/nipy
Once we get it working and clean up the API, we will ask for a code
review before contributing it to scipy.  Clearly, we should have done
this in the first place, but live and learn I guess.  Our hope is that
we will be able to get it in for the scipy 0.8 release.

We would love to have other contributors.  If anyone is interested in
helping, please let me know.

Thanks,

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From wesmckinn at gmail.com  Tue Oct 21 00:34:48 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Tue, 21 Oct 2008 00:34:48 -0400
Subject: [SciPy-user] where is models.py?
In-Reply-To: <c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>
References: <48FD2850.5090308@american.edu>
	<c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>
Message-ID: <30911767-95EC-4418-94CF-38E7434D194A@gmail.com>

What is the timeline for the 0.7 release? I use the models package  
extensively and would personally be interested in developing it  
sooner rather than later if it could potentially be a part of the new  
release.

Thanks,
Wes

On Oct 20, 2008, at 11:02 PM, Jarrod Millman wrote:

> On Mon, Oct 20, 2008 at 5:54 PM, Alan G Isaac <aisaac at american.edu>  
> wrote:
>> models.py seems to have been moved?
>
> Assuming you are referring to the models package, I removed it in  
> August:
> http://projects.scipy.org/scipy/scipy/changeset/4639
> (If not, sorry for the confusion.)
>
> I sent an email to the list before hand explaining that
> scipy.stats.models wasn't ready for release and proposed removing it
> so that we could move forward on the 0.7 release.  I haven't looked at
> the code since then, but as I recall there were several broken tests
> and some open questions about what the API should be.  I spoke with
> Jonathan Taylor (he wrote the package) and he said he wasn't going to
> have time in the near future to work on it.
>
> This is very unfortunate since the statistical models package provides
> some really useful functionality and I would love to see it in the 0.7
> release.  For the time being, it is living (and hopefully developing)
> in the nipy project:  https://launchpad.net/nipy
> Once we get it working and clean up the API, we will ask for a code
> review before contributing it to scipy.  Clearly, we should have done
> this in the first place, but live and learn I guess.  Our hope is that
> we will be able to get it in for the scipy 0.8 release.
>
> We would love to have other contributors.  If anyone is interested in
> helping, please let me know.
>
> Thanks,
>
> -- 
> Jarrod Millman
> Computational Infrastructure for Research Labs
> 10 Giannini Hall, UC Berkeley
> phone: 510.643.4014
> http://cirl.berkeley.edu/
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user



From millman at berkeley.edu  Tue Oct 21 02:26:16 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Mon, 20 Oct 2008 23:26:16 -0700
Subject: [SciPy-user] where is models.py?
In-Reply-To: <30911767-95EC-4418-94CF-38E7434D194A@gmail.com>
References: <48FD2850.5090308@american.edu>
	<c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>
	<30911767-95EC-4418-94CF-38E7434D194A@gmail.com>
Message-ID: <c7009a550810202326s36912e7bv20d370fbe9443348@mail.gmail.com>

On Mon, Oct 20, 2008 at 9:34 PM, Wes McKinney <wesmckinn at gmail.com> wrote:
> What is the timeline for the 0.7 release? I use the models package
> extensively and would personally be interested in developing it
> sooner rather than later if it could potentially be a part of the new
> release.

I would like to have released 0.7, but there are still too many
failing tests.  I just checked everything out and ran the test suite
on a Fedora Linux box and got several test failures (KNOWNFAIL=6,
SKIP=7, errors=6, failures=10; see the attached log for details).  I
am sure that the problems are even worse on the other platforms.
Before we get the release out we are going to have to focus on bugs
squashing for a bit.  So if you are interested in helping get models
back into the trunk, you can try and squash a few bugs as a warm up ;)

Once we get 0.7 released, I will be happy to discuss putting
scipy.models back in the trunk.  (If you are interested in working on
models, but can't help out on other scipy bugs, maybe we could see who
else is interested in working on the models code and try to organize a
sprint at Berkeley.  Please let me know what your interest level/area
is and I can see if I can help.)

I would also like to get build scripts for Windows and Mac finished.
David Cournapeau has been working on the Window's build scripts:
http://projects.scipy.org/scipy/scipy/browser/trunk/tools/win32/build_scripts
I need to check with him to see what the status is (his last check-in
was less than an hour ago).  I believe that the Mac build works fine
and I just need to check them in.  (I am out of town currently, but I
can look into this over the weekend.)

Once I get back to Berkeley, I will try to do a more concerted push to
get everyone to start killing bugs so we can release 0.7 soon.

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tests_scipy.log.gz
Type: application/x-gzip
Size: 4296 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081020/ac109526/attachment.bin>

From david at ar.media.kyoto-u.ac.jp  Tue Oct 21 02:19:58 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 21 Oct 2008 15:19:58 +0900
Subject: [SciPy-user] where is models.py?
In-Reply-To: <30911767-95EC-4418-94CF-38E7434D194A@gmail.com>
References: <48FD2850.5090308@american.edu>	<c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>
	<30911767-95EC-4418-94CF-38E7434D194A@gmail.com>
Message-ID: <48FD748E.1080508@ar.media.kyoto-u.ac.jp>

Wes McKinney wrote:
> What is the timeline for the 0.7 release? I use the models package  
> extensively and would personally be interested in developing it  
> sooner rather than later if it could potentially be a part of the new  
> release.
>
>   

IIRC, the main problem of the models package was its dependency on
weave. I am not familiar with weave much, but I think the consensus was
it was not adapted for scipy packages: it causes some build issues
(since weave itself is in scipy, there is a bootstrap problem for once).
So if you rewrote the part which depends on weave, it could be
reincluded in scipy.

But to be honest, I think we should not accept more code in scipy: scipy
is already quite big, and there is not enough man power. More code means
more maintenance problems. Would it be a big problem for you if the code
was put in a scikit ?

In any case, I am strongly against any new code for scipy 0.7. We are
already really late, there are tons of bugs, and not many people fixing
them. Adding code just make it worse at this point.

cheers,

David


From david at ar.media.kyoto-u.ac.jp  Tue Oct 21 02:32:03 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Tue, 21 Oct 2008 15:32:03 +0900
Subject: [SciPy-user] where is models.py?
In-Reply-To: <c7009a550810202326s36912e7bv20d370fbe9443348@mail.gmail.com>
References: <48FD2850.5090308@american.edu>	<c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>	<30911767-95EC-4418-94CF-38E7434D194A@gmail.com>
	<c7009a550810202326s36912e7bv20d370fbe9443348@mail.gmail.com>
Message-ID: <48FD7763.7090608@ar.media.kyoto-u.ac.jp>

Jarrod Millman wrote:
>
> I would also like to get build scripts for Windows and Mac finished.
> David Cournapeau has been working on the Window's build scripts:
> http://projects.scipy.org/scipy/scipy/browser/trunk/tools/win32/build_scripts
> I need to check with him to see what the status is (his last check-in
> was less than an hour ago). 

I am hoping to get it done for today. The idea was to have it ready so
that we can quickly generate alpha and beta versions for testing on
windows. There is still the problem that the installer is big (around 40
Mb), but I don't see any easy solution in the short term for this,

David


From millman at berkeley.edu  Tue Oct 21 03:28:55 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Tue, 21 Oct 2008 00:28:55 -0700
Subject: [SciPy-user] where is models.py?
In-Reply-To: <48FD748E.1080508@ar.media.kyoto-u.ac.jp>
References: <48FD2850.5090308@american.edu>
	<c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>
	<30911767-95EC-4418-94CF-38E7434D194A@gmail.com>
	<48FD748E.1080508@ar.media.kyoto-u.ac.jp>
Message-ID: <c7009a550810210028h244cb5bbsa43d1c8a1294a580@mail.gmail.com>

On Mon, Oct 20, 2008 at 11:19 PM, David Cournapeau
<david at ar.media.kyoto-u.ac.jp> wrote:
> IIRC, the main problem of the models package was its dependency on
> weave. I am not familiar with weave much, but I think the consensus was
> it was not adapted for scipy packages: it causes some build issues
> (since weave itself is in scipy, there is a bootstrap problem for once).
> So if you rewrote the part which depends on weave, it could be
> reincluded in scipy.

Chris Burns and Tom Waite rewrote the weave bit just before I decided
to pull it:
http://projects.scipy.org/scipy/scipy/changeset/4602
http://projects.scipy.org/scipy/scipy/changeset/4630
http://projects.scipy.org/scipy/scipy/changeset/4631
http://projects.scipy.org/scipy/scipy/changeset/4632

While they were removing the weave dependency, it became apparent that
the packages wasn't ready for release.

> But to be honest, I think we should not accept more code in scipy: scipy
> is already quite big, and there is not enough man power. More code means
> more maintenance problems. Would it be a big problem for you if the code
> was put in a scikit ?

I don't think anyone in my group is going to have time to make it into
a scikit.  Unless there is some serious interest and people willing to
work on it, I expect that we will work on the models code in the nipy
trunk going forward.

> In any case, I am strongly against any new code for scipy 0.7. We are
> already really late, there are tons of bugs, and not many people fixing
> them. Adding code just make it worse at this point.

+1.  Any new code or functionality should be worked on outside of
trunk at least until 0.7 is released.

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From millman at berkeley.edu  Tue Oct 21 03:30:01 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Tue, 21 Oct 2008 00:30:01 -0700
Subject: [SciPy-user] where is models.py?
In-Reply-To: <48FD7763.7090608@ar.media.kyoto-u.ac.jp>
References: <48FD2850.5090308@american.edu>
	<c7009a550810202002r2f346f7dk94d5d9643a76fc48@mail.gmail.com>
	<30911767-95EC-4418-94CF-38E7434D194A@gmail.com>
	<c7009a550810202326s36912e7bv20d370fbe9443348@mail.gmail.com>
	<48FD7763.7090608@ar.media.kyoto-u.ac.jp>
Message-ID: <c7009a550810210030x338d754bv2928552cb6ec6239@mail.gmail.com>

On Mon, Oct 20, 2008 at 11:32 PM, David Cournapeau
<david at ar.media.kyoto-u.ac.jp> wrote:
> I am hoping to get it done for today. The idea was to have it ready so
> that we can quickly generate alpha and beta versions for testing on
> windows. There is still the problem that the installer is big (around 40
> Mb), but I don't see any easy solution in the short term for this,

Excellent.  I really wouldn't worry about the size.  There are a lot
more important things for you to devote your time to at this point.
Thanks for doing this.

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From olli.sipila at helsinki.fi  Tue Oct 21 05:33:12 2008
From: olli.sipila at helsinki.fi (Olli =?utf-8?b?U2lwaWzDpA==?=)
Date: Tue, 21 Oct 2008 12:33:12 +0300
Subject: [SciPy-user] scipy 0.6 build error
In-Reply-To: <48FC7027.9010502@ar.media.kyoto-u.ac.jp>
References: <20081018150641.17457hwrt91wafkh@webmail.helsinki.fi>
	<5b8d13220810191900i7d511b25vc6bb678ef821a97a@mail.gmail.com>
	<20081020095106.58557anc9o1evwwq@webmail.helsinki.fi>
	<48FC7027.9010502@ar.media.kyoto-u.ac.jp>
Message-ID: <20081021123312.52972pjc1sc1x0ig@webmail.helsinki.fi>


I reverted back to 2.5 and managed to build it now. I originally had  
trouble with 2.5, but it seems it had more to do with my account  
permissions than with python itself. Thanks for pointing out the  
problem with 2.6!

- Olli

Quoting "David Cournapeau" <david at ar.media.kyoto-u.ac.jp>:

> Olli Sipil? wrote:
>> Hm ok, I'll give it a go. I recall 2.5 also giving me an error message
>> (this I think is the reason why I wanted to try 2.6), but I'll post
>> that error here as well if 2.5 doesn't work either.
>>
>
> Yes, please report any error on 2.5. Note that even if the error is not
> specific to 2.6, you should not (yet) use numpy and/or scipy with 2.6.
>
> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>




From elmico.filos at gmail.com  Tue Oct 21 05:52:04 2008
From: elmico.filos at gmail.com (=?ISO-8859-1?Q?Mico_Fil=F3s?=)
Date: Tue, 21 Oct 2008 11:52:04 +0200
Subject: [SciPy-user] Simple combinatorics with Numpy
Message-ID: <e700999f0810210252y7e51d2afvcee413928d332c3f@mail.gmail.com>

Hi,

I would like to use NumPy/SciPy to do some basic combinatorics on
small (size<6) 1D arrays of integers.

Imagine I have an array x=([1,3,5,8]) from which I draw, with replacement, a
sample of, say, 3 numbers. The ordering of the sample is unimportant. I want to
calculate the histogram of the averages of the samples (i.e., avg(1,1,1)=1,
avg(1,3,5)=3, etc). For that, I need to *enumerate* all the possible unordered
samples of size 3 that can be drawn from x, i.e.,

  (1,1,1), (1,1,3), (1,1,5), (1,1,8), (1,3,3), (1,5,5), ...
  [there are comb(4+3-1,3)=20 different samples]

and then, I need to compute, for each of them, the weight c/(4**3), where c is a
multinomial factor that takes into account possible repetitions of elements in
the sample.

Although I think I could come up with some dirty, inefficient lines of code
doing the job, I wonder whether there is any NumPy/SciPy function / trick that
could come in handy. I have seen that in Mathematica there is a function called
'Compositions(n,n)' that returns a list with all the possible arrangements of n
nonnegative integers that sum up n (e.g., Compositions(2,2) =
{(0,2),(1,1),(2,0)}). There is also a function 'KSubset' that returns all the
subsets of a given size that can be formed from a larger set. These function
are useful to generate all the possible samples. Is there anything similar in
Numpy/Scipy?

Efficiency is not really an issue here, since the sets are always small. I am
aware that this is not the type of job NumPy is originally designed for, but
since the problem is simple and the implementation seems feasible (in
principle), I wanted to hear the opinion/hints of the experts on that. My
apologies if the answer turns out to be trivial.

Thanks for your attention.


From mnandris at blueyonder.co.uk  Tue Oct 21 14:18:52 2008
From: mnandris at blueyonder.co.uk (Michael)
Date: Tue, 21 Oct 2008 19:18:52 +0100
Subject: [SciPy-user] Simple combinatorics with Numpy
In-Reply-To: <mailman.11.1224608403.19369.scipy-user@scipy.org>
References: <mailman.11.1224608403.19369.scipy-user@scipy.org>
Message-ID: <1224613132.7437.4.camel@mik>

On Tue, 2008-10-21 at 12:00 -0500, scipy-user-request at scipy.org wrote:

>  For that, I need to *enumerate* all the possible unordered
> samples of size 3 that can be drawn from x, i.e.,
> 
>   (1,1,1), (1,1,3), (1,1,5), (1,1,8), (1,3,3), (1,5,5), ...
>   [there are comb(4+3-1,3)=20 different samples]

>>> from scipy.misc import comb
>>> comb(6,3)
array(20.000000000000014)

To get the actual samples, there is a recipe in the Python Cookbook, 2nd
edition, p725

-- 
"When you think of the long and gloomy history of man, you will find far
more hideous crimes have been committed in the name of obedience than
have been committed in the name of rebellion". C.P.Snow,
"Either-Or" (1961)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081021/fec5ed61/attachment.sig>

From tjhnson at gmail.com  Tue Oct 21 17:25:49 2008
From: tjhnson at gmail.com (T J)
Date: Tue, 21 Oct 2008 14:25:49 -0700
Subject: [SciPy-user] Simple combinatorics with Numpy
In-Reply-To: <e700999f0810210252y7e51d2afvcee413928d332c3f@mail.gmail.com>
References: <e700999f0810210252y7e51d2afvcee413928d332c3f@mail.gmail.com>
Message-ID: <b02c60870810211425gd65d1e6pc605a00abd9d0035@mail.gmail.com>

On Tue, Oct 21, 2008 at 2:52 AM, M F wrote:
>
> Imagine I have an array x=([1,3,5,8]) from which I draw, with replacement, a
> sample of, say, 3 numbers. The ordering of the sample is unimportant. I want to
> calculate the histogram of the averages of the samples (i.e., avg(1,1,1)=1,
> avg(1,3,5)=3, etc). For that, I need to *enumerate* all the possible unordered
> samples of size 3 that can be drawn from x, i.e.,

There was a post in numpy-discussion a couple of days ago which
provided a possible solution:

http://projects.scipy.org/pipermail/numpy-discussion/2008-October/038178.html

def boxings(n, k):
    seq, i = [n]*k + [0], k
    while i:
        yield tuple(seq[i] - seq[i+1] for i in xrange(k))
        i = seq.index(0) - 1
        seq[i:k] = [seq[i] - 1] * (k-i)

This enumerates the ways to place n indistinguishable tokens into k
distinguishable boxes (order of boxes is not of concern) when there is
no maximum on the occupancy of any box.  You want to make j unordered
selections from m distinguishable items when the items are replaced
after each selection.

n <-> j
k <-> m

In the example you suggested, boxings(3,4) does the trick, but this
gives you the number of times each item was selected.  There is
probably a simple way to directly generate the samples, but it is
possible to do it using the above function too.

def samples_ur(items, k):
    """Returns k unordered samples (with replacement) from items."""
    n = len(items)
    for sample in boxings(k, n):
        selections = [[items[i]]*count for i,count in enumerate(sample)]
        yield tuple([x for sel in selections for x in sel])


>From there, you should be able to do what you want (eg average).

> and then, I need to compute, for each of them, the weight c/(4**3), where c is a
> multinomial factor that takes into account possible repetitions of elements in
> the sample.

What exactly are you looking for with c?  Do you want to know the
number of permutations for each enumerated item?  This isn't the
prettiest code, but if so, the attached file does what you want.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: samples.py
Type: text/x-python
Size: 1139 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081021/ec29db9d/attachment.py>

From tjhnson at gmail.com  Tue Oct 21 17:26:21 2008
From: tjhnson at gmail.com (T J)
Date: Tue, 21 Oct 2008 14:26:21 -0700
Subject: [SciPy-user] Simple combinatorics with Numpy
In-Reply-To: <1224613132.7437.4.camel@mik>
References: <mailman.11.1224608403.19369.scipy-user@scipy.org>
	<1224613132.7437.4.camel@mik>
Message-ID: <b02c60870810211426i871856cjf2bcec0b919e359@mail.gmail.com>

On Tue, Oct 21, 2008 at 11:18 AM, Michael <mnandris at blueyonder.co.uk> wrote:
> On Tue, 2008-10-21 at 12:00 -0500, scipy-user-request at scipy.org wrote:
>
>>  For that, I need to *enumerate* all the possible unordered
>> samples of size 3 that can be drawn from x, i.e.,
>>
>>   (1,1,1), (1,1,3), (1,1,5), (1,1,8), (1,3,3), (1,5,5), ...
>>   [there are comb(4+3-1,3)=20 different samples]
>
>>>> from scipy.misc import comb
>>>> comb(6,3)
> array(20.000000000000014)
>
> To get the actual samples, there is a recipe in the Python Cookbook, 2nd
> edition, p725
>

I could be wrong, but I don't think the recipe provides what the OP is
requesting. The corresponding docstring would be:

''' take n (not necessarily distinct) items, order is irrelevant '''

which is, noticeably, absent from page 725
(http://books.google.com/books?id=Q0s6Vgb98CQC&pg=PA725)


From dineshbvadhia at hotmail.com  Tue Oct 21 17:42:20 2008
From: dineshbvadhia at hotmail.com (Dinesh B Vadhia)
Date: Tue, 21 Oct 2008 14:42:20 -0700
Subject: [SciPy-user] Calling Scipy Sparse from C++
Message-ID: <COL103-DS25BDB13D56DD08DFAF0159A32E0@phx.gbl>

Hi!  Is it possible to call the Scipy Sparse library (or a Python module that includes Scipy Sparse code) from a C++ program and if so, where can I find some reference information with examples?  Thank-you.

Dinesh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081021/681e8145/attachment.html>

From travis at enthought.com  Tue Oct 21 17:53:51 2008
From: travis at enthought.com (Travis Vaught)
Date: Tue, 21 Oct 2008 16:53:51 -0500
Subject: [SciPy-user] ANN: Enthought Python Distribution - New Release
Message-ID: <02DFC911-B86B-4F92-A83F-1FB7AADAB98B@enthought.com>

Greetings,

Enthought, Inc. is very pleased to announce the newest release of the  
Enthought Python Distribution (EPD) Py2.5 v4.0.30002:

http://www.enthought.com/epd

This release contains updates to many of EPD's packages, including  
NumPy, IPython, matplotlib, VTK, etc.  This is also the first release  
to include a 3.x version of the Enthought Tool Suite (http://code.enthought.com/ 
).

The release notes for this release, including the list of included  
packages, may be found here:
https://svn.enthought.com/epd/wiki/Python2.5.2/4.0.300/GA

Many thanks to the EPD team for putting this release together, and to  
the community of folks who have provided all of the valuable tools  
bundled here.

Best Regards,

Travis


---------
About EPD
---------
The Enthought Python Distribution (EPD) is a "kitchen-sink-included"  
distribution of the Python? Programming Language, including over 80  
additional tools and libraries. The EPD bundle includes NumPy, SciPy,  
IPython, 2D and 3D visualization, database adapters, and a lot of  
other tools right out of the box.

http://www.enthought.com/products/epd.php

It is currently available as an easy, single-click installer for  
Windows XP (x86), Mac OS X (a universal binary for Intel 10.4 and  
above) and RedHat EL3 (x86 and amd64).

EPD is free for 30-day trial use and for use in degree-granting  
academic institutions.  An annual Subscription and installation  
support are available for commercial use (http://www.enthought.com/products/epddownload.php 
) including an Enterprise Subscription with support for particular  
deployment environments (http://www.enthought.com/products/enterprise.php 
).



From wesmckinn at gmail.com  Tue Oct 21 17:57:18 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Tue, 21 Oct 2008 17:57:18 -0400
Subject: [SciPy-user] Calling Scipy Sparse from C++
In-Reply-To: <COL103-DS25BDB13D56DD08DFAF0159A32E0@phx.gbl>
References: <COL103-DS25BDB13D56DD08DFAF0159A32E0@phx.gbl>
Message-ID: <6c476c8a0810211457v2c20387cq142dcf30b90fd753@mail.gmail.com>

Look here:

http://www.python.org/doc/2.5.2/ext/ext.html

I've been successful embedding the Python interpreter in a C++ application--
the best way I found was to create a function that calls a "bridge" module
and translates the output from your C++ program (which has to be a PyObject
*), or alternately you could create a NumPy array inside C++ and pass that
(though I haven't tried this). If you create a generic interface you can
then hot-swap Python code making use of numpy / scipy, much better than
coding C++!

- Wes

On Tue, Oct 21, 2008 at 5:42 PM, Dinesh B Vadhia
<dineshbvadhia at hotmail.com>wrote:

>  Hi!  Is it possible to call the Scipy Sparse library (or a Python module
> that includes Scipy Sparse code) from a C++ program and if so, where can I
> find some reference information with examples?  Thank-you.
>
> Dinesh
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081021/61d0dba1/attachment.html>

From dcday137 at gmail.com  Wed Oct 22 00:42:36 2008
From: dcday137 at gmail.com (Collin Day)
Date: Wed, 22 Oct 2008 04:42:36 +0000 (UTC)
Subject: [SciPy-user]
	=?utf-8?q?Difference_between_ffts=3F_-_plots_of_what?=
	=?utf-8?q?_i_am=09getting=2E?=
References: <20081014174622.128f368b@Krypton.homenet>
	<48F54D82.5000005@ar.media.kyoto-u.ac.jp>
	<loom.20081015T034324-752@post.gmane.org>
Message-ID: <loom.20081022T043925-745@post.gmane.org>



Sorry - just wondering if any of this stuff makes sense?  I suppose I could just
not use the fftpack.fft - it is the one that appears to be "incorrect"






From lionel at gamr7.com  Wed Oct 22 05:18:15 2008
From: lionel at gamr7.com (Lionel Barret De Nazaris)
Date: Wed, 22 Oct 2008 11:18:15 +0200
Subject: [SciPy-user] Arc-length reparametrization. (newbie question ?)
In-Reply-To: <48FC5142.508@gamr7.com>
References: <48FC5142.508@gamr7.com>
Message-ID: <48FEEFD7.8050303@gamr7.com>

Hi,

I answer to myself in the case that someone google this in the future.

Here follows my solution.
This is suboptimal and looks like a naive translation but it does works 
and is short (i.e manageable)

# cp are the curve control points, it is an array

# Bernstein polynomial
# (see http://en.wikipedia.org/wiki/Bernstein_blending_function)
t  = poly1d([1, 0])    # t
s  = poly1d([-1, 1])   # 1-t

# Bezier curve as combination of Bernstein polynomial
# ( see http://en.wikipedia.org/wiki/Bezier_curve)
ps        = [s*s*s, 3*t*s*s, 3*s*t*t, t*t*t]  # position
ps_prime  = map(lambda p: p.deriv(), ps)      # first derivative
ps_second = map(lambda p: p.deriv(m = 2), ps) # second derivative

def Y(t, cp):
    """ the position on the curve at time t """
    ts = array( [p(t) for p in ps] ) # calculating the t coeff
    i  = ts * cp                     # appliying the t coeff to the 
control points
    x = i[0,:].sum()                 # selection the xs and summing them
    y = i[1,:].sum()                 # selection the ys and summing them
    return x, y

def DY(t, cp):
    """ the velocity vector on the curve at time t """
    ts = array( [p(t) for p in ps_prime] )
    i  = ts * cp
    x = i[0,:].sum()                
    y = i[1,:].sum()               
    return x, y

def Speed(t, cp):
    """ the speed (i.e velocity magnitude) on the curve at time t """
    x, y = DY(t, cp)
    return sqrt(x*x+y*y)

def ArcLength(t, cp, tmin=0):
    """ the curve length  corresponding to time t """
    return quad(Speed, tmin, t, args=(cp,))[0]

def getCurveParameter(s, cp, max_tries, epsilon):
    """ the time t corresponding to the curve length s """
    L = ArcLength(1,cp)
    t = 0 + s * (1-0)/L
    for i in range(max_tries):
        F = ArcLength(t, cp) - s
        if abs(F) < epsilon:
            return t
        else:
            DF = Speed(t, cp)
            t -= F/DF
    return t

def getCurveParameter2(s, cp):
    """ the time t corresponding to the curve length s """
    L = ArcLength(1, cp)
    t = 0 + s * (1-0)/L
    f = lambda t : ArcLength(t, cp) - s
    return newton(f, t)

Lionel Barret De Nazaris wrote:
> Hello all,
>
> I've just inherited from a bunch of not so-good code about cubic
> splines, and as a relative scipy newbie, I was wondering what would be 
> the right way to do it.
>
> I need to do Arc-length reparametrization.( 
> http://www.math.hmc.edu/~gu/math142/mellon/Differential_Geometry/Geometry_of_curves/Parametric_Curves_and_arc.html 
> <http://www.math.hmc.edu/%7Egu/math142/mellon/Differential_Geometry/Geometry_of_curves/Parametric_Curves_and_arc.html> 
> )
>
> In the current code, I recognize the use of the newton's method to 
> converge on a good estimate but this leave me non-plussed. The heavy use 
> of integration to compute the curve length without keeping the 
> intermediary results is really weird.
>
> So how would you do it, using the best of what scipy has to offer ?
>
> Note : I've looked as the splines in scipy , but this seems more focused 
> on finding the spline that fits the samples, than on this kind of 
> manipulation.
> Here we have the spline and its control points.
>
> this the code  (see http://pastebin.com/m7e276652 if indentation is not 
> correct) :
>
> <code>
> def derivated_curvilign_abscissa(t, points):
>    return 
> Vector(vectorial_derivated_cubic_interpolator(t,points)).magnitude()
>
> def curvilign_abscissa(t, points):
>    """
>     input :
>       points = spline control points [init_point, init_virtual_point, 
> end_virtual_point, end_point]
>       t      = a float between 0 and 1
>     ouput :
>       the arc-length between f(t) and the start of the curve (aka 
> curvilign_abscissa)
>     """
>    return Vector(points[0]).magnitude() + quad(lambda x 
> :derivated_curvilign_abscissa( x, points), 0., t)[0]
>
> def create_curvy_grid(points, samples):
>    """
>    input :
>       points  = spline control points [init_point, init_virtual_point, 
> end_virtual_point, end_point]
>       samples = list of of equally spaced t (float) between 0 and 1
>    output :
>        a list of float t  (float) where moving from tn to tn+1 means an 
> advancing on the curve of an equal length s.
>    """
>    curve_length = curvilign_abscissa(1., points) -curvilign_abscissa(0., 
> points)
>    # ==
>    def function_to_solve(x, translation) :
>        return - translation + curvilign_abscissa(x, points) 
> -curvilign_abscissa(0., points)
>    # ==
>    return [newton(lambda x : function_to_solve(x, curve_length * 
> step_grid), 1 - PRECISION) for step_grid in grid]
>    # ==
> </code>
>
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
>   

-- 
Best Regards,
lionel Barret de Nazaris,
Gamr7 Founder & CTO
=========================
Gamr7 : Cities for Games
http://www.gamr7.com






From opossumnano at gmail.com  Wed Oct 22 05:04:23 2008
From: opossumnano at gmail.com (Tiziano Zito)
Date: Wed, 22 Oct 2008 11:04:23 +0200
Subject: [SciPy-user] Modular toolkit for Data Processing 2.4 released!
Message-ID: <20081022090423.GA18890@localhost>

We are glad to announce release 2.4 of the Modular toolkit for Data
Processing (MDP).

MDP is a Python library of widely used data processing algorithms that
can be combined according to a pipeline analogy to build more complex
data processing software. The base of available algorithms includes,
to name but the most common, Principal Component Analysis (PCA and
NIPALS), several Independent Component Analysis algorithms (CuBICA,
FastICA, TDSEP, and JADE), Slow Feature Analysis, Restricted Boltzmann
Machine, and Locally Linear Embedding.

What's new in version 2.4?
--------------------------------------

- The new version introduces a new parallel package to execute the MDP
  algorithms on multiple processors or machines. The package also offers
  an interface to develop customized schedulers and parallel algorithms.
  Old MDP scripts can be turned into their parallelized equivalent with
  one simple command.

- The number of available algorithms is increased with the Locally
  Linear Embedding and Hessian eigenmaps algorithms to perform
  dimensionality reduction and manifold learning (many thanks to Jake
  VanderPlas for his contribution!)

- Some more bug fixes, useful features, and code migration towards
  Python 3.0

Resources
---------
Download: http://sourceforge.net/project/showfiles.php?group_id=116959
Homepage: http://mdp-toolkit.sourceforge.net
Mailing list: http://sourceforge.net/mail/?group_id=116959

--

 Pietro Berkes
 Volen Center for Complex Systems
 Brandeis University
 Waltham, MA, USA

 Niko Wilbert
 Institute for Theoretical Biology
 Humboldt-University
 Berlin, Germany

 Tiziano Zito
 Bernstein Center for Computational Neuroscience
 Humboldt-University
 Berlin, Germany



From nwagner at iam.uni-stuttgart.de  Wed Oct 22 05:56:09 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 22 Oct 2008 11:56:09 +0200
Subject: [SciPy-user] Fourier series
Message-ID: <web-113178549@uni-stuttgart.de>

Hi all,
  
Is there a function in scipy to compute the Fourier 
coefficients
a_0, a_1, b_1, a_2, b_2 of a periodic function f(t)=f(t+T)

http://en.wikipedia.org/wiki/Fourier_series

An example would be appreciated.

Nils


From aarchiba at physics.mcgill.ca  Wed Oct 22 06:44:01 2008
From: aarchiba at physics.mcgill.ca (Anne Archibald)
Date: Wed, 22 Oct 2008 06:44:01 -0400
Subject: [SciPy-user] Fourier series
In-Reply-To: <web-113178549@uni-stuttgart.de>
References: <web-113178549@uni-stuttgart.de>
Message-ID: <ce557a360810220344r11354aebr59438a7209fa69dd@mail.gmail.com>

2008/10/22 Nils Wagner <nwagner at iam.uni-stuttgart.de>:

> Is there a function in scipy to compute the Fourier
> coefficients
> a_0, a_1, b_1, a_2, b_2 of a periodic function f(t)=f(t+T)
>
> http://en.wikipedia.org/wiki/Fourier_series
>
> An example would be appreciated.

Not exactly. The usual way to do such a thing would be to evaluate the
function on a grid and use a fast fourier transform:

n = 1024

xs = np.arange(n)*T/float(n)
ys = f(xs)
ft = np.fft.rfft(ys)

You will need to adjust the normalizations somewhat. You will also
need to choose a large enough n to capture all features of interest -
I recommend running it with at least 2 and 4 times the number of
Fourier coefficients you need and comparing. (The extra ones are to
avoid missing features that fall between grid points.)

This method assumes your computational cost would be dominated by the
Fourier transform. If instead your computational cost is dominated by
the function evaluations, you'll have to come up with something more
clever that does adaptive sampling to find all points of interest in
the function. You might try something like using scipy.integrate.quad
and storing the results of all evaluations. You'd then use these
points to evaluate your Fourier transform.

Anne


From david at ar.media.kyoto-u.ac.jp  Wed Oct 22 06:36:04 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 22 Oct 2008 19:36:04 +0900
Subject: [SciPy-user] scipy super pack installer for win32: please test
Message-ID: <48FF0214.4080302@ar.media.kyoto-u.ac.jp>

Hi,

    A quick note to mention I have generated a "superpack" installer for
scipy, for testing purposes. This is similar to numpy superpack
installer: the installer detects your CPU at installation time and
install the right scipy:

http://www.ar.media.kyoto-u.ac.jp/members/david/archives/numpy/scipy-0.7.0.dev4826-win32-superpack-python2.5.exe

How to test:
------------

Inside python:

import scipy
scipy.test()

If python does not crash, this is OK. Test failures are OK, since those
will be fixed within the 0.7.0 release. Please report any problem on
scipy-dev ML,

cheers,

David


From Joris.DeRidder at ster.kuleuven.be  Wed Oct 22 06:54:43 2008
From: Joris.DeRidder at ster.kuleuven.be (Joris De Ridder)
Date: Wed, 22 Oct 2008 12:54:43 +0200
Subject: [SciPy-user] Fourier series
In-Reply-To: <web-113178549@uni-stuttgart.de>
References: <web-113178549@uni-stuttgart.de>
Message-ID: <9947772F-BD46-4ADA-AD18-BB601AFDD61F@ster.kuleuven.be>


On 22 Oct 2008, at 11:56 , Nils Wagner wrote:

> Hi all,
>
> Is there a function in scipy to compute the Fourier
> coefficients
> a_0, a_1, b_1, a_2, b_2 of a periodic function f(t)=f(t+T)
>
> http://en.wikipedia.org/wiki/Fourier_series
>
> An example would be appreciated.
>
> Nils

I guess you mean numerical, not analytical? As far as I know it's only  
possible if you can set an upper limit to its bandwidth. Suppose you  
know for certain that no frequencies higher than 100 Hz occur in your  
signal, than you can uniformly discretize the signal with spacing  
delta t  = 1/(2*100) = 0.005, and then use FFT (http://www.scipy.org/Numpy_Example_List 
) to compute the Fourier coefficients.

Cheers,
Joris


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm



From elmico.filos at gmail.com  Wed Oct 22 06:57:11 2008
From: elmico.filos at gmail.com (=?ISO-8859-1?Q?Mico_Fil=F3s?=)
Date: Wed, 22 Oct 2008 12:57:11 +0200
Subject: [SciPy-user] Simple combinatorics with Numpy
In-Reply-To: <b02c60870810211425gd65d1e6pc605a00abd9d0035@mail.gmail.com>
References: <e700999f0810210252y7e51d2afvcee413928d332c3f@mail.gmail.com>
	<b02c60870810211425gd65d1e6pc605a00abd9d0035@mail.gmail.com>
Message-ID: <e700999f0810220357v533bfdcawbe4a62a4b1076643@mail.gmail.com>

Thanks a lot for your answers.

The functions in sample.py do perfectly the job.

The motivation for this unordered sampling with replacement was perhaps a
little unclear in my first post. I need to count the different ways that a mean
sample of size 3 can occur. The probability of a given value for the sample
mean depends on how many *ordered* samples give rise to that given value. For
instance,

   {1,1,1} = [(1,1,1)]                          -> 1
   {1,1,3} = [(1,1,3), (1,3,1), (3,1,1)]   -> 5/3
   {1,3,5} = [(1,3,5), (3,1,5), (1,5,3), (5,1,3), (5,3,1), (3,5,1)]   -> 8/3
   {1,3,3} = [(1,3,3), (3,1,3), (3,3,1)]   -> 7/3
   {3,3,3} = [(3,3,3)]                     -> 3 ...

where {} and () denote the unordered and ordered samples, respectively. The
number of different orderings associated with a particular sample is given by a
multinomial distribution: if there are 3 places and 4 different numbers
repeated k1, k2, k3, k4 times, there are c = 3!/(k1!k2!k3!k4!) ordered samples.

Of course, I could also have proceeded the brute force way, enumerating all 3^4
ordered samples, computing their average, and counting the number of times each
different value occurs.

Sorry for being so wordy.


From david at ar.media.kyoto-u.ac.jp  Wed Oct 22 07:59:57 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Wed, 22 Oct 2008 20:59:57 +0900
Subject: [SciPy-user] Fourier series
In-Reply-To: <9947772F-BD46-4ADA-AD18-BB601AFDD61F@ster.kuleuven.be>
References: <web-113178549@uni-stuttgart.de>
	<9947772F-BD46-4ADA-AD18-BB601AFDD61F@ster.kuleuven.be>
Message-ID: <48FF15BD.1040402@ar.media.kyoto-u.ac.jp>

Joris De Ridder wrote:
>
> I guess you mean numerical, not analytical? As far as I know it's only  
> possible if you can set an upper limit to its bandwidth.

Concretely, by having only some samples of your functions, there is an
implied bandwidth for your signal (otherwise, the sampling process would
have destroyed your signal through aliasing). Then, FFT coefficients can
approach the Fourier coefficients by refining the frequency "sampling"
(considering the FFT as a frequency sampled of the Fourier coefficients).

Also, numerically speaking, you can consider that any real signal has
finite bandwidth, or more exactly the coefficients are negligeable for
high frequency: the fourier coefficients are decreasing toward 0 for
smooth functions.  More precisely, the Fourier transform TF(n) "behaves
as" 1/n^k where k is bigger when the function is more regular: the sum
of (1+n^-k)^2 * F(n) is finite for the function k-times differentiable.
Incidentally, one mathematical object for the study of smooth functions
is Sobolev space, which definition is based on this property.

cheers,

David


From ivo.maljevic at gmail.com  Wed Oct 22 10:05:53 2008
From: ivo.maljevic at gmail.com (Ivo Maljevic)
Date: Wed, 22 Oct 2008 10:05:53 -0400
Subject: [SciPy-user] Fourier series
In-Reply-To: <web-113178549@uni-stuttgart.de>
References: <web-113178549@uni-stuttgart.de>
Message-ID: <826c64da0810220705g228156d8s89cf4bf2a7c56b8a@mail.gmail.com>

I think if you spent some time in deriving formulas it should be possible to
find it:

1) Since the function f(t) is periodic, its spectrum is discrete anyway, so
FFT will do the job, provided that you sample the frequency at multiple of
harmonics.

2) DFT coefficients are complex, but similar to c_n complex coefficients for
the fourier series, there is a relationship:
c_n = a_n -j*b_n, n > 0
c_n = a_n+j*b_n, n < 0

So, for positive m: a_m = 0.5*(c_m + c_{-m}), and similar approach goes for
b_m.

Sorry I cannot give you more info at this time. Hopefully this sketchy
approach will give you some ideas.

Ivo



2008/10/22 Nils Wagner <nwagner at iam.uni-stuttgart.de>

> Hi all,
>
> Is there a function in scipy to compute the Fourier
> coefficients
> a_0, a_1, b_1, a_2, b_2 of a periodic function f(t)=f(t+T)
>
> http://en.wikipedia.org/wiki/Fourier_series
>
> An example would be appreciated.
>
> Nils
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081022/5ed26fd3/attachment.html>

From josef.pktd at gmail.com  Wed Oct 22 11:22:06 2008
From: josef.pktd at gmail.com (joep)
Date: Wed, 22 Oct 2008 08:22:06 -0700 (PDT)
Subject: [SciPy-user] scipy super pack installer for win32: please test
In-Reply-To: <48FF0214.4080302@ar.media.kyoto-u.ac.jp>
References: <48FF0214.4080302@ar.media.kyoto-u.ac.jp>
Message-ID: <5ed2aa23-d992-40d8-90c5-eb34968ffa2a@k37g2000hsf.googlegroups.com>


> http://www.ar.media.kyoto-u.ac.jp/members/david/archives/numpy/scipy-0.7.0.dev4826-win32-superpack-python2.5.exe

installs without problems on windowsXP, sse2
The only problem I had was that I had a trunk version of scipy (on a
non site-packages directory) in easy-install.pth, which I needed to
comment out. Initially it loaded this version instead of the one
installed
in site-packages by the superpack.

One weird thing I found, when I run scipy.test() in a command shell,
or ipython shell, then I get a large output with Depreciation Warnings
and test output. When I run scipy.test() in an Idle shell, I don't get
any Depreciation Warnings or test output. I have no idea why the
output should be different.

Another observation which is not clear but is not very important is
that, when I compiled the trunk version a few weeks ago (with MingW)
and I run scipy.test(), then I don't get the 2 test failures in
test_lapack.test_all_lapack. (my lapack and atlas files were
downloaded a long time ago from the scipy.org install description
page)

Josef


Python 2.5.2 (r252:60911, Feb 21 2008, 13:11:45) [MSC v.1310 32 bit
(Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import scipy
>>> scipy.__file__
'C:\\Programs\\Python25\\lib\\site-packages\\scipy\\__init__.pyc'
>>> scipy.test()
Running unit tests for scipy
NumPy version 1.2.0rc2
NumPy is installed in C:\Programs\Python25\lib\site-packages\numpy
SciPy version 0.7.0.dev4826
SciPy is installed in C:\Programs\Python25\lib\site-packages\scipy
Python version 2.5.2 (r252:60911, Feb 21 2008, 13:11:45) [MSC v.1310
32 bit (Int
el)]
nose version 0.10.4

...
----------------------------------------------------------------------
Ran 2317 tests in 46.969s

FAILED (KNOWNFAIL=2, SKIP=14, failures=2)
<nose.result.TextTestResult run=2317 errors=0 failures=2>
>>>
>>> scipy.show_config()
umfpack_info:
  NOT AVAILABLE

dfftw_info:
  NOT AVAILABLE

blas_opt_info:
    libraries = ['f77blas', 'cblas', 'atlas']
    library_dirs = ['C:\\local\\lib\\yop\\sse2']
    define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
    language = c

mkl_info:
  NOT AVAILABLE

djbfft_info:
  NOT AVAILABLE

atlas_blas_threads_info:
  NOT AVAILABLE

lapack_opt_info:
    libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
    library_dirs = ['C:\\local\\lib\\yop\\sse2']
    define_macros = [('ATLAS_INFO', '"\\"?.?.?\\""')]
    language = f77

fftw2_info:
  NOT AVAILABLE

fftw3_info:
  NOT AVAILABLE

atlas_info:
    libraries = ['lapack', 'f77blas', 'cblas', 'atlas']
    library_dirs = ['C:\\local\\lib\\yop\\sse2']
    language = f77

lapack_mkl_info:
  NOT AVAILABLE

blas_mkl_info:
  NOT AVAILABLE

atlas_blas_info:
    libraries = ['f77blas', 'cblas', 'atlas']
    library_dirs = ['C:\\local\\lib\\yop\\sse2']
    language = c

atlas_threads_info:
  NOT AVAILABLE

>>>

2 failures are

======================================================================
FAIL: test_lapack.test_all_lapack
----------------------------------------------------------------------
Traceback (most recent call last):
  File "c:\programs\python25\lib\site-packages\nose-0.10.4-py2.5.egg
\nose\case.p
y", line 182, in runTest
    self.test(*self.arg)
  File "C:\Programs\Python25\Lib\site-packages\scipy\lib\lapack\tests
\esv_tests.
py", line 41, in check_syevr
    assert_array_almost_equal(w,exact_w)
  File "C:\Programs\Python25\Lib\site-packages\numpy\testing
\utils.py", line 311
, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "C:\Programs\Python25\Lib\site-packages\numpy\testing
\utils.py", line 296
, in assert_array_compare
    raise AssertionError(msg)
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])

======================================================================
FAIL: test_lapack.test_all_lapack
----------------------------------------------------------------------
Traceback (most recent call last):
  File "c:\programs\python25\lib\site-packages\nose-0.10.4-py2.5.egg
\nose\case.p
y", line 182, in runTest
    self.test(*self.arg)
  File "C:\Programs\Python25\Lib\site-packages\scipy\lib\lapack\tests
\esv_tests.
py", line 66, in check_syevr_irange
    assert_array_almost_equal(w,exact_w[rslice])
  File "C:\Programs\Python25\Lib\site-packages\numpy\testing
\utils.py", line 311
, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File "C:\Programs\Python25\Lib\site-packages\numpy\testing
\utils.py", line 296
, in assert_array_compare
    raise AssertionError(msg)
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])

































From david at ar.media.kyoto-u.ac.jp  Wed Oct 22 11:27:11 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Thu, 23 Oct 2008 00:27:11 +0900
Subject: [SciPy-user] scipy super pack installer for win32: please test
In-Reply-To: <5ed2aa23-d992-40d8-90c5-eb34968ffa2a@k37g2000hsf.googlegroups.com>
References: <48FF0214.4080302@ar.media.kyoto-u.ac.jp>
	<5ed2aa23-d992-40d8-90c5-eb34968ffa2a@k37g2000hsf.googlegroups.com>
Message-ID: <48FF464F.9090906@ar.media.kyoto-u.ac.jp>

joep wrote:
>
> installs without problems on windowsXP, sse2
> The only problem I had was that I had a trunk version of scipy (on a
> non site-packages directory) in easy-install.pth, which I needed to
> comment out. Initially it loaded this version instead of the one
> installed
> in site-packages by the superpack.

Unfortunately, there is not much I can do here, since this is a
setuptools problem. In my experience, the easy-install.pth is fragile,
and when you start using it, you should be ready to edit it by hand
occasionally.

>
> One weird thing I found, when I run scipy.test() in a command shell,
> or ipython shell, then I get a large output with Depreciation Warnings
> and test output. When I run scipy.test() in an Idle shell, I don't get
> any Depreciation Warnings or test output. I have no idea why the
> output should be different.

Maybe something linked to the Idle shell: warning and deprecation can be
controlled by the python interpreter. I find it a bit strange to ignore
warnings by default, though.

>
> Another observation which is not clear but is not very important is
> that, when I compiled the trunk version a few weeks ago (with MingW)
> and I run scipy.test(), then I don't get the 2 test failures in
> test_lapack.test_all_lapack. (my lapack and atlas files were
> downloaded a long time ago from the scipy.org install description
> page)

Different flags/compilers/atlas sources can cause some differences in
the results. The errors in the tests are small but not negligeable; I
don't know if they are significant. The superpack uses ATLAS 3.8.2, and
it does not produce those errors on linux; so this may be caused by
relatively old compilers on windows (mingw gcc is really old if you use
the released ones). To complicate the matter, a given ATLAS build is not
reproducible, making those issues even more difficult to track.

Thanks for testing,

David


From ivo.maljevic at gmail.com  Wed Oct 22 13:32:32 2008
From: ivo.maljevic at gmail.com (Ivo Maljevic)
Date: Wed, 22 Oct 2008 13:32:32 -0400
Subject: [SciPy-user] Fourier series
In-Reply-To: <web-113178549@uni-stuttgart.de>
References: <web-113178549@uni-stuttgart.de>
Message-ID: <826c64da0810221032mee82227k8e4296fcc2ce448e@mail.gmail.com>

Maybe this simple matlab code can help you to develop ideas further:

clear
N = 1000;

%time covers only one period, excluding the last point, as that
% is the begining of the next period
t=(0:N-1)/N;
dt = t(2) - t(1);
% set the sampling frequency to some reasonable value
f=1/dt/5

% uncomment if you want only one cos function. This function is periodic,
% period T = 1 / f => fundamental frequency is
%x=cos(2*pi*f*t);
x=cos(2*pi*f*t)+cos(4*pi*f*t);
X = 1/length(t)*fft(x);

nonzero_index = find(abs(X) > 1e-3);
l = length(nonzero_index);
n = (nonzero_index(1:l/2)-1)/f
X(nonzero_index)

a = 2*real(X(nonzero_index))
b = -2*imag(X(nonzero_index))

As you know, cos() is a periodic function with only one non-zero coefficient
in the fourier sereis. In this trivial example,
you will get n =1 and a_n and b_n coefficients if you use the signal from
the first line, or n=1,2 for the signal from the second line.
It works even if you lower N down to 10 or some other number that is a
multiple of 5 (for this example).

Look at the comments too, to know when to end the period of the signal.

Granted, this is a simplified example, and you may want to figure out how
many sample points you need. Just to remind you,
from Parseval's theorem you know that the power of a periodic signal (which
has a finite power) is the sum of all these
coefficients squared. This means that for any signal that is F-transformable
you will find that you only need a finite number
of sample points.

This is of course far away from the real deal, I just thougt that this might
give you some ideas on which way to go. Or, maybe somebody else has already
done this, which I wouldn't be surprised, and you will get your answer
without much work.

Ivo

 . In thi


2008/10/22 Nils Wagner <nwagner at iam.uni-stuttgart.de>

> Hi all,
>
> Is there a function in scipy to compute the Fourier
> coefficients
> a_0, a_1, b_1, a_2, b_2 of a periodic function f(t)=f(t+T)
>
> http://en.wikipedia.org/wiki/Fourier_series
>
> An example would be appreciated.
>
> Nils
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081022/a256e06a/attachment.html>

From erik at erikwickstrom.com  Wed Oct 22 13:55:25 2008
From: erik at erikwickstrom.com (Erik Wickstrom)
Date: Wed, 22 Oct 2008 10:55:25 -0700
Subject: [SciPy-user] Some help with chisquare
Message-ID: <3d381e170810221055q262d308ah1377a0f0f0fc6c7d@mail.gmail.com>

Hi,

I'm trying to port an application to python, and want to use scipy to handle
the statistics.

The app takes several tests and uses chi-square to determines which has the
highest success rate with a confidence of 95% or better (critical
values/degrees of freedom).

For example:
Test a:
Total trials = 100
Total successes = 60

Test b:
Total trials = 105
Total successes = 46

Test c:
Total trials = 98
Total successes = 52

It then puts the data through some sort of chi-square formula (or so the
comments say) and produces a chi-square value that can be compared against
the critical values for 95% confidence.

Trouble is, I'm not sure which of the many scipy chi-square functions to
use, and what data I need to feed into them....

Any suggestions?
Thanks!
Erik
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081022/82d44e4f/attachment.html>

From reckoner at gmail.com  Wed Oct 22 14:36:43 2008
From: reckoner at gmail.com (Reckoner)
Date: Wed, 22 Oct 2008 11:36:43 -0700
Subject: [SciPy-user] 2008 scipy conference review
Message-ID: <a1b53c580810221136t4e585003o51a210b075e2ff37@mail.gmail.com>

the following link is a review of the 2008 scipy conference in Pasadena:

https://www.osc.edu/cms/sip/node/4


From borreguero at gmail.com  Wed Oct 22 17:22:34 2008
From: borreguero at gmail.com (Jose Borreguero)
Date: Wed, 22 Oct 2008 17:22:34 -0400
Subject: [SciPy-user] How to write to a NetCDF file with 64-bit offset mode?
Message-ID: <7cced4ed0810221422w12960981pd9a0b991eaefbe36@mail.gmail.com>

Dear scipy users,

I open a NetCDF file for writing with:
  object=NetCDFFile(outfile, 'w')

I need to create a file with the 64-bit offset mode on. Is there a flag like
NC_64BIT_OFFSET that I can pass to the constructor? If so, then how?

regards,
-- 
Jose M. Borreguero
Postdoctoral Associate
Oak Ridge National Laboratory
P.O. Box 2008, M.S. 6164
Oak Ridge, TN 37831
phone: 865-241-3071 fax: 865-576-5491
Email: borreguerojm at ornl.gov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081022/51565ee0/attachment.html>

From aisaac at american.edu  Wed Oct 22 18:04:42 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Wed, 22 Oct 2008 18:04:42 -0400
Subject: [SciPy-user] scipy super pack installer for win32: please test
In-Reply-To: <5ed2aa23-d992-40d8-90c5-eb34968ffa2a@k37g2000hsf.googlegroups.com>
References: <48FF0214.4080302@ar.media.kyoto-u.ac.jp>
	<5ed2aa23-d992-40d8-90c5-eb34968ffa2a@k37g2000hsf.googlegroups.com>
Message-ID: <48FFA37A.4050804@american.edu>

>> http://www.ar.media.kyoto-u.ac.jp/members/david/archives/numpy/scipy-0.7.0.dev4826-win32-superpack-python2.5.exe 

joep wrote:
> when I run scipy.test() in a command shell, 
> or ipython shell, then I get a large output with Depreciation Warnings 
> and test output. 

Similar experience. Three failures. (I know these are not
a current concern.)  Output below.
Alan


Python 2.5.2 (r252:60911, Feb 21 2008, 13:11:45) [MSC v.1310
32 bit (Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys, scipy
>>> sys.stderr = open('c:/temp/temp.out','w')
>>> scipy.test()
Running unit tests for scipy
NumPy version 1.2.0
NumPy is installed in C:\Python25\lib\site-packages\numpy
SciPy version 0.7.0.dev4826
SciPy is installed in C:\Python25\lib\site-packages\scipy
Python version 2.5.2 (r252:60911, Feb 21 2008, 13:11:45) [MSC v.1310 32 
bit (Int
el)]
nose version 0.10.1
Warning: 1000000 bytes requested, 20 bytes read.
caxpy:n=4
caxpy:n=3
ccopy:n=4
ccopy:n=3
cscal:n=4
cswap:n=4
cswap:n=3
daxpy:n=4
daxpy:n=3
dcopy:n=4
dcopy:n=3
dscal:n=4
dswap:n=4
dswap:n=3
saxpy:n=4
saxpy:n=3
scopy:n=4
scopy:n=3
sscal:n=4
sswap:n=4
sswap:n=3
zaxpy:n=4
zaxpy:n=3
zcopy:n=4
zcopy:n=3
zscal:n=4
zswap:n=4
zswap:n=3
ATLAS version 3.8.2 built by david on Tue Aug  5 13:01:25 TST 2008:
    UNAME    : CYGWIN_NT-5.1 donau-win 1.5.25(0.156/4/2) 2008-06-12 
19:34 i686 Cy
gwin
    INSTFLG  : -1 0 -a 1
    ARCHDEFS : -DATL_OS_WinNT -DATL_ARCH_P4 -DATL_CPUMHZ=3200 -DGCCWIN 
-DUseClock
  -DATL_SSE2 -DATL_SSE1 -DATL_GAS_x8632
    F2CDEFS  : -DAdd__ -DF77_INTEGER=int -DStringSunStyle
    CACHEEDGE: 3145728
    F77      : g77, version GNU Fortran (GCC) 3.4.4 (cygming special, 
gdc 0.12, u
sing dmd 0.125)
    F77FLAGS : -O -m32
    SMC      : gcc, version gcc (GCC) 3.4.4 (cygming special, gdc 0.12, 
using dmd
  0.125)
    SMCFLAGS : -fomit-frame-pointer -mfpmath=387 -O2 -falign-loops=4 -m32
    SKC      : gcc, version gcc (GCC) 3.4.4 (cygming special, gdc 0.12, 
using dmd
  0.125)
    SKCFLAGS : -fomit-frame-pointer -mfpmath=387 -O2 -falign-loops=4 -m32
caxpy:n=4
caxpy:n=3
ccopy:n=4
ccopy:n=3
cscal:n=4
cswap:n=4
cswap:n=3
daxpy:n=4
daxpy:n=3
dcopy:n=4
dcopy:n=3
dscal:n=4
dswap:n=4
dswap:n=3
saxpy:n=4
saxpy:n=3
scopy:n=4
scopy:n=3
sscal:n=4
sswap:n=4
sswap:n=3
zaxpy:n=4
zaxpy:n=3
zcopy:n=4
zcopy:n=3
zscal:n=4
zswap:n=4
zswap:n=3
Result may be inaccurate, approximate err = 9.82300567882e-009
Result may be inaccurate, approximate err = 1.87387729512e-010
warning: specified build_dir '_bad_path_' does not exist or is not 
writable. Try
ing default locations
warning: specified build_dir '_bad_path_' does not exist or is not 
writable. Try
ing default locations
error removing c:\docume~1\alanis~1\locals~1\temp\tmpkbpafhcat_test: 
c:\docume~1
\alanis~1\locals~1\temp\tmpkbpafhcat_test: The directory is not empty
building extensions here: c:\docume~1\alanis~1\locals~1\temp\Alan 
Isaac\python25
_compiled\m4
<nose.result.TextTestResult run=2317 errors=0 failures=3>
>>>


And here is the error log:

C:\Python25\lib\site-packages\scipy\linsolve\__init__.py:4:
DeprecationWarning: scipy.linsolve has moved to
scipy.sparse.linalg.dsolve
   warn('scipy.linsolve has moved to scipy.sparse.linalg.dsolve', 
DeprecationWarning)
.................................................................................................................................C:\Python25\lib\site-packages\scipy\cluster\vq.py:570: 
UserWarning: One of the clusters is empty. Re-run kmean with a different 
initialization.
   warnings.warn("One of the clusters is empty. "
...................................................C:\Python25\lib\site-packages\scipy\integrate\odepack.py:144: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   ixpr, mxstep, mxhnil, mxordn, mxords)
C:\Python25\lib\site-packages\scipy\integrate\odepack.py:144: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   ixpr, mxstep, mxhnil, mxordn, mxords)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:304: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return 
_quadpack._qawse(func,a,b,wvar,integr,args,full_output,epsabs,epsrel,limit)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:304: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return 
_quadpack._qawse(func,a,b,wvar,integr,args,full_output,epsabs,epsrel,limit)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:306: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return 
_quadpack._qawce(func,a,b,wvar,args,full_output,epsabs,epsrel,limit)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:306: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return 
_quadpack._qawce(func,a,b,wvar,args,full_output,epsabs,epsrel,limit)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:295: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return 
_quadpack._qawfe(thefunc,-b,wvar,integr,args,full_output,epsabs,limlst,limit,maxp1)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:295: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return 
_quadpack._qawfe(thefunc,-b,wvar,integr,args,full_output,epsabs,limlst,limit,maxp1)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:249: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return _quadpack._qagse(func,a,b,args,full_output,epsabs,epsrel,limit)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:249: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return _quadpack._qagse(func,a,b,args,full_output,epsabs,epsrel,limit)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:251: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return 
_quadpack._qagie(func,bound,infbounds,args,full_output,epsabs,epsrel,limit)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:251: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return 
_quadpack._qagie(func,bound,infbounds,args,full_output,epsabs,epsrel,limit)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:273: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return 
_quadpack._qawoe(func,a,b,wvar,integr,args,full_output,epsabs,epsrel,limit,maxp1,1)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:273: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return 
_quadpack._qawoe(func,a,b,wvar,integr,args,full_output,epsabs,epsrel,limit,maxp1,1)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:280: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return 
_quadpack._qawfe(func,a,wvar,integr,args,full_output,epsabs,limlst,limit,maxp1)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:280: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return 
_quadpack._qawfe(func,a,wvar,integr,args,full_output,epsabs,limlst,limit,maxp1)
.C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:259: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   return 
_quadpack._qagpe(func,a,b,the_points,args,full_output,epsabs,epsrel,limit)
C:\Python25\lib\site-packages\scipy\integrate\quadpack.py:259: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   return 
_quadpack._qagpe(func,a,b,the_points,args,full_output,epsabs,epsrel,limit)
......C:\Python25\lib\site-packages\scipy\interpolate\fitpack2.py:479: 
UserWarning:
The coefficients of the spline returned have been computed as the
minimal norm least-squares solution of a (numerically) rank deficient
system (deficiency=7). If deficiency is large, the results may be
inaccurate. Deficiency may strongly depend on the value of eps.
   warnings.warn(message)
....C:\Python25\lib\site-packages\scipy\interpolate\fitpack2.py:420: 
UserWarning:
The required storage space exceeds the available storage space: nxest
or nyest too small, or s too small.
The weighted least-squares spline corresponds to the current set of
knots.
   warnings.warn(message)
...........C:\Python25\lib\site-packages\scipy\interpolate\fitpack.py:760: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   tx,ty,nxest,nyest,wrk,lwrk1,lwrk2)
C:\Python25\lib\site-packages\scipy\interpolate\fitpack.py:760: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   tx,ty,nxest,nyest,wrk,lwrk1,lwrk2)
C:\Python25\lib\site-packages\scipy\interpolate\fitpack.py:837: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   z,ier=_fitpack._bispev(tx,ty,c,kx,ky,x,y,dx,dy)
C:\Python25\lib\site-packages\scipy\interpolate\fitpack.py:837: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   z,ier=_fitpack._bispev(tx,ty,c,kx,ky,x,y,dx,dy)
...............................C:\Python25\lib\site-packages\scipy\interpolate\fitpack.py:485: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   y,ier=_fitpack._spl_(x,der,t,c,k)
C:\Python25\lib\site-packages\scipy\interpolate\fitpack.py:485: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   y,ier=_fitpack._spl_(x,der,t,c,k)
..................KK..C:\Python25\lib\site-packages\scipy\io\tests\test_array_import.py:29: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   b = numpyio.fread(fid,1000000,N.Int16,N.Int)
C:\Python25\lib\site-packages\scipy\io\tests\test_array_import.py:29: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   b = numpyio.fread(fid,1000000,N.Int16,N.Int)
.C:\Python25\lib\site-packages\numpy\lib\utils.py:106: 
DeprecationWarning: write_array is deprecated
   warnings.warn(str1, DeprecationWarning)
C:\Python25\lib\site-packages\numpy\lib\utils.py:106: 
DeprecationWarning: read_array is deprecated
   warnings.warn(str1, DeprecationWarning)
......................C:\Python25\lib\site-packages\numpy\lib\utils.py:106: 
DeprecationWarning: npfile is deprecated
   warnings.warn(str1, DeprecationWarning)
....................................................................................................................................................................................FF................................................................................................................................................................................................................................................................................................F...........................................................................................................................................................................................................................................................................................................................................................................................................C:\Python25\lib\site-packages\scipy\odr\odrpack.py:1055: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   self.output = Output(apply(odr, args, kwds))
C:\Python25\lib\site-packages\scipy\odr\odrpack.py:1055: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   self.output = Output(apply(odr, args, kwds))
.................C:\Python25\lib\site-packages\scipy\optimize\minpack.py:270: 
DeprecationWarning: PyArray_FromDims: use PyArray_SimpleNew.
   retval = 
_minpack._lmdif(func,x0,args,full_output,ftol,xtol,gtol,maxfev,epsfcn,factor,diag)
C:\Python25\lib\site-packages\scipy\optimize\minpack.py:270: 
DeprecationWarning: PyArray_FromDimsAndDataAndDescr: use 
PyArray_NewFromDescr.
   retval = 
_minpack._lmdif(func,x0,args,full_output,ftol,xtol,gtol,maxfev,epsfcn,factor,diag)
.................................SSSSSSSSSSS...............................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................C:\Python25\lib\site-packages\scipy\stats\stats.py:413: 
DeprecationWarning: scipy.stats.mean is deprecated; please update your 
code to use numpy.mean.
Please note that:
     - numpy.mean axis argument defaults to None, not 0
     - numpy.mean has a ddof argument to replace bias in a more general 
manner.
       scipy.stats.mean(a, bias=True) can be replaced by numpy.mean(x,
axis=0, ddof=1).
   axis=0, ddof=1).""", DeprecationWarning)
.C:\Python25\lib\site-packages\scipy\stats\stats.py:1237: 
DeprecationWarning: scipy.stats.std is deprecated; please update your 
code to use numpy.std.
Please note that:
     - numpy.std axis argument defaults to None, not 0
     - numpy.std has a ddof argument to replace bias in a more general 
manner.
       scipy.stats.std(a, bias=True) can be replaced by numpy.std(x,
axis=0, ddof=1).
   axis=0, ddof=1).""", DeprecationWarning)
C:\Python25\lib\site-packages\scipy\stats\stats.py:1214: 
DeprecationWarning: scipy.stats.var is deprecated; please update your 
code to use numpy.var.
Please note that:
     - numpy.var axis argument defaults to None, not 0
     - numpy.var has a ddof argument to replace bias in a more general 
manner.
       scipy.stats.var(a, bias=True) can be replaced by numpy.var(x,
axis=0, ddof=1).
   axis=0, ddof=1).""", DeprecationWarning)
.C:\Python25\lib\site-packages\scipy\stats\morestats.py:618: 
UserWarning: Ties preclude use of exact statistic.
   warnings.warn("Ties preclude use of exact statistic.")
......C:\Python25\lib\site-packages\scipy\stats\stats.py:491: 
DeprecationWarning: scipy.stats.median is deprecated; please update your 
code to use numpy.median.
Please note that:
     - numpy.median axis argument defaults to None, not 0
     - numpy.median has a ddof argument to replace bias in a more 
general manner.
       scipy.stats.median(a, bias=True) can be replaced by numpy.median(x,
axis=0, ddof=1).
   axis=0, ddof=1).""", DeprecationWarning)
......................................................C:\Python25\lib\site-packages\numpy\lib\function_base.py:343: 
Warning:
             The semantics of histogram has been modified in
             the current release to fix long-standing issues with
             outliers handling. The main changes concern
             1. the definition of the bin edges,
                now including the rightmost edge, and
             2. the handling of upper outliers, now ignored rather
                than tallied in the rightmost bin.
             The previous behaviour is still accessible using
             `new=False`, but is scheduled to be deprecated in the
             next release (1.3).

             *This warning will not printed in the 1.3 release.*

             Use `new=True` to bypass this warning.

             Please read the docstring for more information.

   """, Warning)
....................................................................................................................................................................................................................................
======================================================================
FAIL: test_lapack.test_all_lapack
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\lib\site-packages\nose\case.py", line 203, in runTest
     self.test(*self.arg)
   File 
"C:\Python25\Lib\site-packages\scipy\lib\lapack\tests\esv_tests.py", 
line 41, in check_syevr
     assert_array_almost_equal(w,exact_w)
   File "C:\Python25\Lib\site-packages\numpy\testing\utils.py", line 
311, in assert_array_almost_equal
     header='Arrays are not almost equal')
   File "C:\Python25\Lib\site-packages\numpy\testing\utils.py", line 
296, in assert_array_compare
     raise AssertionError(msg)
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
  x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
  y: array([-0.66992434,  0.48769389,  9.18223045])

======================================================================
FAIL: test_lapack.test_all_lapack
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\lib\site-packages\nose\case.py", line 203, in runTest
     self.test(*self.arg)
   File 
"C:\Python25\Lib\site-packages\scipy\lib\lapack\tests\esv_tests.py", 
line 66, in check_syevr_irange
     assert_array_almost_equal(w,exact_w[rslice])
   File "C:\Python25\Lib\site-packages\numpy\testing\utils.py", line 
311, in assert_array_almost_equal
     header='Arrays are not almost equal')
   File "C:\Python25\Lib\site-packages\numpy\testing\utils.py", line 
296, in assert_array_compare
     raise AssertionError(msg)
AssertionError:
Arrays are not almost equal

(mismatch 33.3333333333%)
  x: array([-0.66992444,  0.48769444,  9.18222618], dtype=float32)
  y: array([-0.66992434,  0.48769389,  9.18223045])

======================================================================
FAIL: test_imresize (test_pilutil.TestPILUtil)
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 82, in skipper
     return f(*args, **kwargs)
   File 
"C:\Python25\Lib\site-packages\scipy\misc\tests\test_pilutil.py", line 
24, in test_imresize
     assert_equal(im1.shape,(11,22))
   File "C:\Python25\Lib\site-packages\numpy\testing\utils.py", line 
174, in assert_equal
     assert_equal(len(actual),len(desired),err_msg,verbose)
   File "C:\Python25\Lib\site-packages\numpy\testing\utils.py", line 
183, in assert_equal
     raise AssertionError(msg)
AssertionError:
Items are not equal:
  ACTUAL: 0
  DESIRED: 2

======================================================================
KNOWNFAIL: test_mio.test_load
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\lib\site-packages\nose\case.py", line 203, in runTest
     self.test(*self.arg)
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 119, in skipper
     raise KnownFailureTest, msg
KnownFailureTest: Test skipped due to known failure

======================================================================
KNOWNFAIL: test_mio.test_round_trip
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\lib\site-packages\nose\case.py", line 203, in runTest
     self.test(*self.arg)
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 119, in skipper
     raise KnownFailureTest, msg
KnownFailureTest: Test skipped due to known failure

======================================================================
SKIP: Getting factors of complex matrix
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Getting factors of real matrix
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Getting factors of complex matrix
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Getting factors of real matrix
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Prefactorize (with UMFPACK) matrix for solving with multiple rhs
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Prefactorize matrix for solving with multiple rhs
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Solve with UMFPACK: double precision complex
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Solve: single precision complex
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Solve with UMFPACK: double precision, sparse rhs
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Solve with UMFPACK: double precision
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

======================================================================
SKIP: Solve: single precision
----------------------------------------------------------------------
Traceback (most recent call last):
   File "C:\Python25\Lib\site-packages\numpy\testing\decorators.py", 
line 80, in skipper
     raise nose.SkipTest, msg
SkipTest: UMFPACK appears not to be compiled

----------------------------------------------------------------------
Ran 2317 tests in 70.461s

FAILED (failures=3)
























From aisaac at american.edu  Wed Oct 22 20:07:58 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Wed, 22 Oct 2008 20:07:58 -0400
Subject: [SciPy-user] scipy super pack installer for win32: please test
In-Reply-To: <48FFA37A.4050804@american.edu>
References: <48FF0214.4080302@ar.media.kyoto-u.ac.jp>	<5ed2aa23-d992-40d8-90c5-eb34968ffa2a@k37g2000hsf.googlegroups.com>
	<48FFA37A.4050804@american.edu>
Message-ID: <48FFC05E.5000208@american.edu>

>>> http://www.ar.media.kyoto-u.ac.jp/members/david/archives/numpy/scipy-0.7.0.dev4826-win32-superpack-python2.5.exe 

> joep wrote:
>> when I run scipy.test() in a command shell, 
>> or ipython shell, then I get a large output with Depreciation Warnings 
>> and test output. 


On 10/22/2008 6:04 PM Alan G Isaac apparently wrote:
> Similar experience. Three failures. (I know these are not
> a current concern.)  Output below.


Hmm, on a different machine (Intel T2600) I get
only 2 errors.  Don't know if that's useful info...
Alan



From dwf at cs.toronto.edu  Thu Oct 23 03:20:30 2008
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 23 Oct 2008 03:20:30 -0400
Subject: [SciPy-user] ndimage starting points
Message-ID: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>

Hi all,

Lately I've been looking at ndimage for replacing some of the  
functionality in the Matlab Image Processing Toolbox and elsewhere,  
but am running into some documentation holes. The parts of ndimage  
that I can figure out how to use work brilliantly and as advertised  
(the filters module for example), but a lot of the functions in some  
submodules don't say much about what form of input they take.

I'm hoping that someone more familiar with the codebase can point me  
in the right direction, and I'll be happy to clean up whatever comes  
of this thread (and anything else I discover) so that we can put it  
into docstrings or the cookbook. I'm mainly focusing on  
ndimage.measurements for now.

So, here's the list. Any help or clarification is appreciated.

- I've figured out, without much help from the docstrings, that the  
labels= argument to many of the functions is an integer array (that  
can be) produced using the very handy label() function. This probably  
deserves a mention in the module docstring (which I am happy to write).

- label() takes an optional "structure" argument - what exactly is  
this, what form does it take, how does one create it, and in what  
circumstances should it be used? Also, is it intended to be used with  
thresholded images?

- The same question about 'structure' goes for watershed_ift, as well  
as what form the 'markers' argument takes (I'm assuming an array with  
markers.shape == input.shape). dtype is... anything numeric I guess?  
It says negatives are treated differently than positives, but nothing  
else.

- center_of_mass() - this may be a dumb question, but this produces an  
"index" (which is not integer valued) in ndim(input) space, where  
higher values in position (i_1, i_2, ... i_n) produce more "pull" on  
the center of mass than a lower value in the same position would?

- Is there a reason that find_objects() takes a "max_label" argument  
whereas every other function takes a scalar or sequence "index"  
argument? It seems inconsistent, though there may be some good  
algorithmic reason for it.

- On a similar note, find_objects computes a "bounding box" of some  
sort when generating slices, I'm guessing? (or are slices far more  
general than I had thought?)

- histogram()'s documentation seems incomplete to me. Just to be  
clear, does it always produce a one-dimensional object, regardless of  
the dimensionality of the input?

- Some things like variance() don't immediately seem to add anything  
to the standard numpy functions, I am assuming that the ability to  
mask by label is their key advantage. Can someone confirm or correct  
this?

That's about all I've got for now.

Thanks in advance,

David


From stefan at sun.ac.za  Thu Oct 23 03:39:28 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Thu, 23 Oct 2008 09:39:28 +0200
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
Message-ID: <9457e7c80810230039l16749388mbcfd72abab05e054@mail.gmail.com>

Hi David

2008/10/23 David Warde-Farley <dwf at cs.toronto.edu>:
> Lately I've been looking at ndimage for replacing some of the
> functionality in the Matlab Image Processing Toolbox and elsewhere,
> but am running into some documentation holes. The parts of ndimage
> that I can figure out how to use work brilliantly and as advertised
> (the filters module for example), but a lot of the functions in some
> submodules don't say much about what form of input they take.

The ndimage module comes from Numeric.  Unfortunately, the version we
included came from a different source and did not have all the Numeric
documentation.  It would be great if you could write a patch to bring
some of the docs over -- they may also answer your questions below.

> - Some things like variance() don't immediately seem to add anything
> to the standard numpy functions, I am assuming that the ability to
> mask by label is their key advantage. Can someone confirm or correct
> this?

Much of the ndimage functionality was implemented before NumPy.  I
think we can already replace a big part of its functionality using
Python + Numpy, without going down to the C level (and I think this
would benefit the library in general).

Cheers
St?fan


From millman at berkeley.edu  Thu Oct 23 03:41:51 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Thu, 23 Oct 2008 00:41:51 -0700
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
Message-ID: <c7009a550810230041s115ad74exa03217d47f2c73ad@mail.gmail.com>

Since you are offering to help out with this, I would like to see
someone do the following:

Take the existing numarray.ndimage docs:
  http://stsdas.stsci.edu/numarray/numarray-1.5.html/module-numarray.ndimage.html
and the cookbook stuff:
  http://www.scipy.org/SciPyPackages/Ndimage
and merge what you can into the docstrings and convert the rest into
restructured text and commit it to the scipy trunk.  That way we can
start working on generating sphinx documentation for scipy.

Thanks,

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From dwf at cs.toronto.edu  Thu Oct 23 04:02:21 2008
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 23 Oct 2008 04:02:21 -0400
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <c7009a550810230041s115ad74exa03217d47f2c73ad@mail.gmail.com>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
	<c7009a550810230041s115ad74exa03217d47f2c73ad@mail.gmail.com>
Message-ID: <72FAC71F-2575-407B-96B2-4869BB320ED5@cs.toronto.edu>

On 23-Oct-08, at 3:41 AM, Jarrod Millman wrote:

> Since you are offering to help out with this, I would like to see
> someone do the following:
>
> Take the existing numarray.ndimage docs:
>  http://stsdas.stsci.edu/numarray/numarray-1.5.html/module-numarray.ndimage.html
> and the cookbook stuff:
>  http://www.scipy.org/SciPyPackages/Ndimage
> and merge what you can into the docstrings and convert the rest into
> restructured text and commit it to the scipy trunk.  That way we can
> start working on generating sphinx documentation for scipy.


Thanks for the link. I don't know how I missed the numarray stuff.

I'd be happy to bring as much of this as possible into docstrings and  
other SciPy docs - I'll keep a few editor windows open as I sort  
through reading them.

David




From dwf at cs.toronto.edu  Thu Oct 23 04:08:35 2008
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Thu, 23 Oct 2008 04:08:35 -0400
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <9457e7c80810230039l16749388mbcfd72abab05e054@mail.gmail.com>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
	<9457e7c80810230039l16749388mbcfd72abab05e054@mail.gmail.com>
Message-ID: <75B1A618-5A46-4170-B01E-E21E15AF5662@cs.toronto.edu>

Hi Stefan,

Thanks for the quick reply!

On 23-Oct-08, at 3:39 AM, St?fan van der Walt wrote:

> Hi David
>
> 2008/10/23 David Warde-Farley <dwf at cs.toronto.edu>:
>
> The ndimage module comes from Numeric.  Unfortunately, the version we
> included came from a different source and did not have all the Numeric
> documentation.  It would be great if you could write a patch to bring
> some of the docs over -- they may also answer your questions below.

Did you mean to say numarray? Jarrod pointed me at the docs for a  
numarray nd_image module. Hopefully there isn't also a module from  
Numeric as well...

Come to think of it, I do know why I missed those docs -- Google isn't  
smart enough to return results for "nd_image" when you search for  
ndimage :)

>> - Some things like variance() don't immediately seem to add anything
>> to the standard numpy functions, I am assuming that the ability to
>> mask by label is their key advantage. Can someone confirm or correct
>> this?
>
> Much of the ndimage functionality was implemented before NumPy.  I
> think we can already replace a big part of its functionality using
> Python + Numpy, without going down to the C level (and I think this
> would benefit the library in general).

So, the one thing that, for example variance() does is take a 'labels'  
parameter and an 'index' parameter, which seems to suggest to me that  
the purpose of this is to compute variance within a labeled object (or  
all labeled objects, disregarding the background). I imagine this  
would be easily reimplemented in pure Python, though, you're right.

For now I'll see about getting this documentation up to snuff using  
the magical tome Jarrod pointed me toward.

Cheers,

David

From stefan at sun.ac.za  Thu Oct 23 04:53:44 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Thu, 23 Oct 2008 10:53:44 +0200
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <75B1A618-5A46-4170-B01E-E21E15AF5662@cs.toronto.edu>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
	<9457e7c80810230039l16749388mbcfd72abab05e054@mail.gmail.com>
	<75B1A618-5A46-4170-B01E-E21E15AF5662@cs.toronto.edu>
Message-ID: <9457e7c80810230153q48bd61a9sdeb92ae8a38d8911@mail.gmail.com>

2008/10/23 David Warde-Farley <dwf at cs.toronto.edu>:
> Did you mean to say numarray? Jarrod pointed me at the docs for a

Yup, sorry.

> For now I'll see about getting this documentation up to snuff using
> the magical tome Jarrod pointed me toward.

Thanks, that'd be much appreciated!

Cheers
St?fan


From macrozhu at gmail.com  Thu Oct 23 08:34:42 2008
From: macrozhu at gmail.com (Macro Zhu)
Date: Thu, 23 Oct 2008 14:34:42 +0200
Subject: [SciPy-user] longfloat print out problem in Windows XP
Message-ID: <11b97ec0810230534w546e6034q422a13530f12fdb1@mail.gmail.com>

Hi,

I met this print out problem for longfloat type:

> print scipy.longfloat(10)
-1.49166814624e-154

> print float(scipy.longfloat(10))
10.0

> print scipy.array(10, dtype='longfloat')
-1.49166814624e-154

The same problem happens with Numpy as well.
How can I get the print out of longfloat correctly, except by converting it
back to the float type?
I am using scipy 0.6.0 and numpy 1.2.0 with python 2.4 in Windows XP SP3.

Thanks!

-Mac
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081023/6ce37b09/attachment.html>

From walter at aims.ac.za  Thu Oct 23 08:28:40 2008
From: walter at aims.ac.za (Walter Mudzimbabwe)
Date: Thu, 23 Oct 2008 14:28:40 +0200 (SAST)
Subject: [SciPy-user] prime numbers
Message-ID: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>


can anybody help me figure out why the following program cannot produce
primes upto 10.
--------------------------------------------------
from scipy import *

def isdivisible(n,listt):
   for i in range(len(listt)):
       if (n%listt[i]==0):
           return 1
       else:
           return 0

def primes_upto(m):
   u=[1,2]
   for i in range(3,m+1):
        if (isdivisible(i,u[1:])==0):
           u.append(i)
   return u

print primes_upto(10)
-----------------------------------------------------
it's output is:

[1, 2, 3, 5, 7, 9]



-- 
Walter Mudzimbabwe (Formerly with AIMS)
University of Western Cape.
Mathematics Dept,
Private Bag X17,
7535 Bellville,
RSA

Contact :+27 78 5188402
         mudzmudz at gmail.com

"Those of many tricks take them to the grave".......waltermudz20008






From david at ar.media.kyoto-u.ac.jp  Thu Oct 23 09:10:34 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Thu, 23 Oct 2008 22:10:34 +0900
Subject: [SciPy-user] longfloat print out problem in Windows XP
In-Reply-To: <11b97ec0810230534w546e6034q422a13530f12fdb1@mail.gmail.com>
References: <11b97ec0810230534w546e6034q422a13530f12fdb1@mail.gmail.com>
Message-ID: <490077CA.2040802@ar.media.kyoto-u.ac.jp>

Macro Zhu wrote:
> Hi,
>
> I met this print out problem for longfloat type:
>
> > print scipy.longfloat(10)
> -1.49166814624e-154
>
> > print float(scipy.longfloat(10))
> 10.0
>
> > print scipy.array(10, dtype='longfloat')
> -1.49166814624e-154
>
> The same problem happens with Numpy as well.
> How can I get the print out of longfloat correctly, except by
> converting it back to the float type?

You can't. I won't bore you with details, but basically, the Microsoft C
runtime (printf) does not handle long double. There is a solution, but
it was too involved when 1.2.0 was about to be released. I hope to solve
the problem for 1.3.0.

Note that the problem is only in the print statement. For example:

a = scipy.longfloat(10)
b = scipy.longfloat(10)
print a # garbage
print a.astype(scipy.float) # 10
print (a+b).astype(scipy.float) # 20

cheers,

David


From gael.varoquaux at normalesup.org  Thu Oct 23 09:34:11 2008
From: gael.varoquaux at normalesup.org (Gael Varoquaux)
Date: Thu, 23 Oct 2008 15:34:11 +0200
Subject: [SciPy-user] prime numbers
In-Reply-To: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
References: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
Message-ID: <20081023133411.GA10344@phare.normalesup.org>

On Thu, Oct 23, 2008 at 02:28:40PM +0200, Walter Mudzimbabwe wrote:

> can anybody help me figure out why the following program cannot produce
> primes upto 10.

I haven't spent more the a few seconds looking at that (little time), but
my guess is:

> --------------------------------------------------
> from scipy import *

> def isdivisible(n,listt):
>    for i in range(len(listt)):
>        if (n%listt[i]==0):
>            return 1
>        else:
>            return 0

"else: ..." should be indented as the "for", not as the "if".

Ga?l



From josef.pktd at gmail.com  Thu Oct 23 09:51:14 2008
From: josef.pktd at gmail.com (joep)
Date: Thu, 23 Oct 2008 06:51:14 -0700 (PDT)
Subject: [SciPy-user] Problems with Google groups
Message-ID: <53ee39e7-4d94-40e9-8575-dbe56fe2c1bd@b1g2000hsg.googlegroups.com>

Just an observation

I tried to reply to the prime numbers question through Google groups,
but the message is declared as deleted or expired. It seems that
Google groups is not very reliable anymore. Numpy-discussion has
disappeared from Google groups a few days ago.

Time for me to look for an alternative.

Josef


From josef.pktd at gmail.com  Thu Oct 23 09:53:24 2008
From: josef.pktd at gmail.com (joep)
Date: Thu, 23 Oct 2008 06:53:24 -0700 (PDT)
Subject: [SciPy-user] prime numbers
In-Reply-To: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
References: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
Message-ID: <1644ab8c-38ff-40b6-87af-cd3ad09d74ac@b1g2000hsg.googlegroups.com>



On Oct 23, 8:28?am, "Walter Mudzimbabwe" <wal... at aims.ac.za> wrote:
> can anybody help me figure out why the following program cannot produce
> primes upto 10.
> --------------------------------------------------
> from scipy import *
>
> def isdivisible(n,listt):
> ? ?for i in range(len(listt)):
> ? ? ? ?if (n%listt[i]==0):
> ? ? ? ? ? ?return 1
> ? ? ? ?else:
> ? ? ? ? ? ?return 0
>
> def primes_upto(m):
> ? ?u=[1,2]
> ? ?for i in range(3,m+1):
> ? ? ? ? if (isdivisible(i,u[1:])==0):
> ? ? ? ? ? ?u.append(i)
> ? ?return u
>
> print primes_upto(10)
> -----------------------------------------------------
> it's output is:
>
> [1, 2, 3, 5, 7, 9]
>
> --
> Walter Mudzimbabwe (Formerly with AIMS)
> University of Western Cape.
> Mathematics Dept,
> Private Bag X17,
> 7535 Bellville,
> RSA
>
> Contact :+27 78 5188402
> ? ? ? ? ?mudzm... at gmail.com
>
> "Those of many tricks take them to the grave".......waltermudz20008
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-u... at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user

return 0 at the end of the for loop

def isdivisible(n,listt):
   for i in range(len(listt)):
       if (n%listt[i]==0):
           return 1
   return 0


From et.gaudrain at free.fr  Thu Oct 23 09:40:53 2008
From: et.gaudrain at free.fr (Etienne Gaudrain)
Date: Thu, 23 Oct 2008 14:40:53 +0100
Subject: [SciPy-user] prime numbers
In-Reply-To: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
References: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
Message-ID: <49007EE5.5090902@free.fr>

Hi,

This is probably not the best place to post your question...
However, this version should work:

--------------------------------------------------
from scipy import *

def isdivisible(n,listt):
   for i in range(len(listt)):
       if (n%listt[i]==0):
           return 1
   return 0

def primes_upto(m):
   u=[1,2]
   for i in range(3,m+1):
        if (isdivisible(i,u[1:])==0):
           u.append(i)
   return u

print primes_upto(10)
-----------------------------------------------------

Note that this is not the best algorithm. Google "primes sieve".

-Etienne




Walter Mudzimbabwe wrote:
> can anybody help me figure out why the following program cannot produce
> primes upto 10.
> --------------------------------------------------
> from scipy import *
>
> def isdivisible(n,listt):
>    for i in range(len(listt)):
>        if (n%listt[i]==0):
>            return 1
>        else:
>            return 0
>
> def primes_upto(m):
>    u=[1,2]
>    for i in range(3,m+1):
>         if (isdivisible(i,u[1:])==0):
>            u.append(i)
>    return u
>
> print primes_upto(10)
> -----------------------------------------------------
> it's output is:
>
> [1, 2, 3, 5, 7, 9]
>
>
>
>   


-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Etienne Gaudrain
Centre for the Neural Basis of Hearing
Department of Physiology, Development and Neuroscience
University of Cambridge
Downing Street
Cambridge CB2 3EG
UK
Phone: +44 (1223) 333 859 office
Fax: +44 (1223) 333 840 department
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~



From ajvogel at tuks.co.za  Thu Oct 23 09:57:53 2008
From: ajvogel at tuks.co.za (Adolph J. Vogel)
Date: Thu, 23 Oct 2008 15:57:53 +0200
Subject: [SciPy-user] prime numbers
In-Reply-To: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
References: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
Message-ID: <200810231557.53357.ajvogel@tuks.co.za>

On Thursday 23 October 2008 14:28:40 Walter Mudzimbabwe wrote:
> can anybody help me figure out why the following program cannot produce
> primes upto 10.
> --------------------------------------------------
> from scipy import *
>
> def isdivisible(n,listt):
>    for i in range(len(listt)):
>        if (n%listt[i]==0):
>            return 1
>        else:
>            return 0
>
> def primes_upto(m):
>    u=[1,2]
>    for i in range(3,m+1):
At this point your only passing the last element of u, which in this case is 
two.  So in isdivisibile() your only checking to see if it is divisible by 
two. You need to check and see whether or not its divisible by the divisors 
smaller then two as well.

Your code is also very complicated for such an simple task.

def isdivisible(p):
    for d in range(1,p):
	if p % d == 0:
		return False
    return True

primes = []
def prime_upto(m):
    for i in range(3,m+1,2):
	if isdivisible(i): primes.append(i)



-- 
Adolph J. Vogel      BEng(Mech)



From josef.pktd at gmail.com  Thu Oct 23 10:02:13 2008
From: josef.pktd at gmail.com (joep)
Date: Thu, 23 Oct 2008 07:02:13 -0700 (PDT)
Subject: [SciPy-user] Problems with Google groups
In-Reply-To: <53ee39e7-4d94-40e9-8575-dbe56fe2c1bd@b1g2000hsg.googlegroups.com>
References: <53ee39e7-4d94-40e9-8575-dbe56fe2c1bd@b1g2000hsg.googlegroups.com>
Message-ID: <ff5a6c3a-e191-4590-a32b-d68163e9a73f@64g2000hsm.googlegroups.com>

Looks like it was just a temporary problem with the prime numbers
thread.

On Oct 23, 9:51?am, joep <josef.p... at gmail.com> wrote:
> Just an observation
>
> I tried to reply to the prime numbers question through Google groups,
> but the message is declared as deleted or expired. It seems that
> Google groups is not very reliable anymore. Numpy-discussion has
> disappeared from Google groups a few days ago.
>
> Time for me to look for an alternative.
>
> Josef
> _______________________________________________
> SciPy-user mailing list
> SciPy-u... at scipy.orghttp://projects.scipy.org/mailman/listinfo/scipy-user


From nwagner at iam.uni-stuttgart.de  Thu Oct 23 10:02:25 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Thu, 23 Oct 2008 16:02:25 +0200
Subject: [SciPy-user] array manipulation
Message-ID: <web-113238708@uni-stuttgart.de>

Hi all,

>>> M
array([[1025338,       1,       1,       1],
        [1036103,       1,       1,       1],
        [2008297,       1,       1,       1],
        [2086888,       0,       0,       1],
        [2127079,       1,       0,       0],
        [2157100,       0,       1,       0],
        [2157969,       1,       1,       1],
        [2222852,       1,       0,       1]])
>>> 1025338 in M[:,0]
True
>>> 2157100 in M[:,0]
True

How can I obtain the row index of numbers that belong
to M ?

Nils


From robert.kern at gmail.com  Thu Oct 23 10:24:55 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 23 Oct 2008 09:24:55 -0500
Subject: [SciPy-user] array manipulation
In-Reply-To: <web-113238708@uni-stuttgart.de>
References: <web-113238708@uni-stuttgart.de>
Message-ID: <3d375d730810230724i7c64efcci4998f8509a73bf5e@mail.gmail.com>

On Thu, Oct 23, 2008 at 09:02, Nils Wagner <nwagner at iam.uni-stuttgart.de> wrote:
> Hi all,
>
>>>> M
> array([[1025338,       1,       1,       1],
>        [1036103,       1,       1,       1],
>        [2008297,       1,       1,       1],
>        [2086888,       0,       0,       1],
>        [2127079,       1,       0,       0],
>        [2157100,       0,       1,       0],
>        [2157969,       1,       1,       1],
>        [2222852,       1,       0,       1]])
>>>> 1025338 in M[:,0]
> True
>>>> 2157100 in M[:,0]
> True
>
> How can I obtain the row index of numbers that belong
> to M ?

searchsorted() if you can guarantee that the column is always sorted.
Otherwise nonzero(x == M[:,0])[0].

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From aisaac at american.edu  Thu Oct 23 10:37:43 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Thu, 23 Oct 2008 10:37:43 -0400
Subject: [SciPy-user] prime numbers
In-Reply-To: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
References: <60401.196.11.235.119.1224764920.squirrel@webmail.aims.ac.za>
Message-ID: <49008C37.8080100@american.edu>

Walter Mudzimbabwe wrote:
> can anybody help me figure out why the following program cannot produce
> primes upto 10.

I think this has nothing to do with SciPy?
Try comp.lang.python instead.

But you can try:

def isdivisible(n, listt):
     return not all(n%d for d in listt)

This will short circuit appropriately.

Alan Isaac


From daniele at grinta.net  Thu Oct 23 13:24:24 2008
From: daniele at grinta.net (Daniele Nicolodi)
Date: Thu, 23 Oct 2008 19:24:24 +0200
Subject: [SciPy-user] Suggestion about algorithm
Message-ID: <4900B348.9050304@grinta.net>

Hello, i'm going to ask something not strictly related to scipy. Forgive 
me if this is not appropriate on the mailing list, but i don't know 
where else i can seek for help, any suggestion is appreciated.

I'm measuring the quality factor Q of a mechanical oscillator. I use the 
ring down technique: i excite the oscillator to a big oscillation 
amplitude so that my read out noise is negligible and then i observe the 
  decay of the oscillation amplitude during time.

The evolution of the amplitude A(t) in time can be described, negletting 
any external perturbation, as:

A(t) = A0 * exp(-t/Beta)

where Q = w0 / 2*Beta and w0 is the oscillator natural frequency.

I usually analyze my data extracting the amplitude of each oscillation 
and then computing:

Beta = - dA(t)/dt / A(t)

where dA(t)/dt is the first derivative of the amplitude computed as the 
difference between the amplitude of the current cicle and the previous 
cicle divided by the period of oscillation.

The problem arises because my oscillator has a very long period (about 
500 seconds) and a very high Q (about 600000). This means that the 
observation time is much shorter than the characteristic time of the 
system and that the value of Beta i want to resolve is very small.
In this situation my uncertainty on Beta is too big to resolve Q.

Does someone have a suggestion for a better technique to analyze my 
data? There is any smarter thing i can do?

Thanks. Bye.
-- 
Daniele


From robert.kern at gmail.com  Thu Oct 23 14:20:47 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Thu, 23 Oct 2008 13:20:47 -0500
Subject: [SciPy-user] Suggestion about algorithm
In-Reply-To: <4900B348.9050304@grinta.net>
References: <4900B348.9050304@grinta.net>
Message-ID: <3d375d730810231120n97f301cs7b3200904f11a615@mail.gmail.com>

On Thu, Oct 23, 2008 at 12:24, Daniele Nicolodi <daniele at grinta.net> wrote:
> Hello, i'm going to ask something not strictly related to scipy. Forgive
> me if this is not appropriate on the mailing list, but i don't know
> where else i can seek for help, any suggestion is appreciated.
>
> I'm measuring the quality factor Q of a mechanical oscillator. I use the
> ring down technique: i excite the oscillator to a big oscillation
> amplitude so that my read out noise is negligible and then i observe the
>  decay of the oscillation amplitude during time.
>
> The evolution of the amplitude A(t) in time can be described, negletting
> any external perturbation, as:
>
> A(t) = A0 * exp(-t/Beta)
>
> where Q = w0 / 2*Beta and w0 is the oscillator natural frequency.
>
> I usually analyze my data extracting the amplitude of each oscillation
> and then computing:
>
> Beta = - dA(t)/dt / A(t)
>
> where dA(t)/dt is the first derivative of the amplitude computed as the
> difference between the amplitude of the current cicle and the previous
> cicle divided by the period of oscillation.
>
> The problem arises because my oscillator has a very long period (about
> 500 seconds) and a very high Q (about 600000). This means that the
> observation time is much shorter than the characteristic time of the
> system and that the value of Beta i want to resolve is very small.
> In this situation my uncertainty on Beta is too big to resolve Q.
>
> Does someone have a suggestion for a better technique to analyze my
> data? There is any smarter thing i can do?

Can you just get the oscillating curve itself rather than extracting
the peaks? It might be easiest just to fit the decaying oscillator
function to the curve. Your uncertainly may still be large, but
probably better than what you currently have.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From millman at berkeley.edu  Thu Oct 23 15:50:21 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Thu, 23 Oct 2008 12:50:21 -0700
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <72FAC71F-2575-407B-96B2-4869BB320ED5@cs.toronto.edu>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
	<c7009a550810230041s115ad74exa03217d47f2c73ad@mail.gmail.com>
	<72FAC71F-2575-407B-96B2-4869BB320ED5@cs.toronto.edu>
Message-ID: <c7009a550810231250n70eb4744t45093e637a948afc@mail.gmail.com>

On Thu, Oct 23, 2008 at 1:02 AM, David Warde-Farley <dwf at cs.toronto.edu> wrote:
> Thanks for the link. I don't know how I missed the numarray stuff.

The nd_image documentation was written in latex.  I don't have time to
look for it right now, but later this week I should be able to get you
the latex sources (unless no one beats me to it).  You could use
pandoc to try and auto-convert them to restructured text for the first
pass.  You will probably need to clean it up a bit, but it will get
you most of the way there.

http://johnmacfarlane.net/pandoc/

Thanks,

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From spmcinerney at hotmail.com  Thu Oct 23 16:53:59 2008
From: spmcinerney at hotmail.com (Stephen McInerney)
Date: Thu, 23 Oct 2008 13:53:59 -0700
Subject: [SciPy-user] prime numbers
In-Reply-To: <mailman.1055.1224768326.2878.scipy-user@scipy.org>
References: <mailman.1055.1224768326.2878.scipy-user@scipy.org>
Message-ID: <BAY111-W1E09A46DD5ED10CAE0423A0280@phx.gbl>


Walter, 
 
The culprit is the else: return 0 clause in your isdivisible() fn
You return 0 immediately if n is not divisible by the first entry listt[0]
i.e. 2. That's why you see all the odd numbers.
 
> def isdivisible(n,listt):> for i in range(len(listt)):> if (n%listt[i]==0):> return 1> else:> return 0
but you want this, where the 'return 0' is a fallthrough when all iterations of the for loop have failed to find any divisor:
> if (n%listt[i]==0):> return 1>
> return 0 # this is a fallthrough 
 
 
PS If you don't omit 1 from listt then you don't need to say listt[1:]
 
PPS another refinement to make this a proper Sieve of Eratosthenes
is that you only need to test up to sqrt(n) rounded down, i.e. int(math.sqrt(n))
Thus e.g. when you consider primeness of 31 you don't need to test if it's
divisible by all primes up to 29, only up to 5. This will speed things up a lot.
 
Regards,
Stephen
 
> Message: 9> Date: Thu, 23 Oct 2008 14:28:40 +0200 (SAST)> From: "Walter Mudzimbabwe" <walter at aims.ac.za>> Subject: [SciPy-user] prime numbers> To: scipy-user at scipy.org> Message-ID:> <60401.196.11.235.119.1224764920.squirrel at webmail.aims.ac.za>> Content-Type: text/plain;charset=iso-8859-1> > > can anybody help me figure out why the following program cannot produce> primes upto 10.> --------------------------------------------------> from scipy import *> > def isdivisible(n,listt):> for i in range(len(listt)):> if (n%listt[i]==0):> return 1> else:> return 0> > def primes_upto(m):> u=[1,2]> for i in range(3,m+1):> if (isdivisible(i,u[1:])==0):> u.append(i)> return u> > print primes_upto(10)> -----------------------------------------------------> it's output is:> > [1, 2, 3, 5, 7, 9]> > > > -- > Walter Mudzimbabwe (Formerly with AIMS)> University of Western Cape.> Mathematics Dept,> Private Bag X17,> 7535 Bellville,> RSA> > Contact :+27 78 5188402> mudzmudz at gmail.com
_________________________________________________________________
Want to read Hotmail messages in Outlook? The Wordsmiths show you how.
http://windowslive.com/connect/post/wedowindowslive.spaces.live.com-Blog-cns!20EE04FBC541789!167.entry?ocid=TXT_TAGLM_WL_hotmail_092008
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081023/ed6d870b/attachment.html>

From stefan at sun.ac.za  Thu Oct 23 17:22:41 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Thu, 23 Oct 2008 23:22:41 +0200
Subject: [SciPy-user] SciPy Sprint Weekend: 1 and 2 November
Message-ID: <9457e7c80810231422m6cab2e75wbcb343472449312@mail.gmail.com>

Event: SciPy Sprint
Date: 1, 2 November 2007
Find us on: irc.freenode.net in channel #scipy

Hi all,

As industry and academia slows down towards the end of the year, we
may be able to tap some developer resources for a final sprint or two
towards SciPy 0.7.  There has never been so much low-hanging fruit
before: the bug tracker is filled with tickets just waiting to be
triaged and fixed!

I would like to hold a SciPy sprint on the 1st and 2nd of November.
At the moment, I can only volunteer my own time, but I am confident
that many of our students and colleagues would be able to lend a hand.
 If some of you are willing to donate 2 hours of your time, this would
be a great opportunity to assist us in getting ready for the next Big
Release of SciPy.

We shall start working on Saturday morning here in South Africa, and
soon thereafter America should wake up to the smell of freshly triaged
tickets.

I look forward to seeing you there, and to the 0.7 release of SciPy!

Kind regards,
St?fan

P.S. I sent this to the scipy-user list, since I firmly believe that
many of our users have the potential to make valuable contributions as
developers.  Don't be shy -- we appreciate help on all levels.

P.S.S. If you are in South Africa, I'll provide free drinks and
internet.  That's almost reason enough to come!


From pav at iki.fi  Thu Oct 23 17:54:17 2008
From: pav at iki.fi (Pauli Virtanen)
Date: Thu, 23 Oct 2008 21:54:17 +0000 (UTC)
Subject: [SciPy-user] ndimage starting points
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
	<c7009a550810230041s115ad74exa03217d47f2c73ad@mail.gmail.com>
Message-ID: <gdqrq9$arj$1@ger.gmane.org>

Thu, 23 Oct 2008 00:41:51 -0700, Jarrod Millman wrote:

> Since you are offering to help out with this, I would like to see
> someone do the following:
> 
> Take the existing numarray.ndimage docs:
>   http://stsdas.stsci.edu/numarray/numarray-1.5.html/module-
numarray.ndimage.html
>
> and the cookbook stuff:
>
>   http://www.scipy.org/SciPyPackages/Ndimage
>
> and merge what you can into the docstrings and convert the rest into
> restructured text and commit it to the scipy trunk.  That way we can
> start working on generating sphinx documentation for scipy.

I started working on Sphinx stuff for Scipy last week, and got this far:

	http://www.iki.fi/pav/tmp/scipy-refguide.tar.gz (source)
	http://www.iki.fi/pav/tmp/scipy-refguide/

It's a Sphinx framework similar to the Numpy reference guide we started 
working on this summer. There's in principle also a BZR branch for this 
in Launchpad,

	https://code.launchpad.net/~pauli-virtanen/scipy/scipy-refguide

which you could track using Bazaar (http://bazaar-vcs.org/), but for some 
reason Launchpad decided to dislike me today, so it doesn't work now.
(BTW, should I put this to Scipy SVN, and if yes, where? What about the 
corresponding Numpy documentation?)

There's already a very simple bare-bones version of the ndimage module 
page there already, but it only lists the functions and their docstrings 
which Sphinx extracts from scipy.ndimage.

-- 
Pauli Virtanen



From sebastian.rooks at free.fr  Thu Oct 23 18:13:43 2008
From: sebastian.rooks at free.fr (Sebastian Rooks)
Date: Thu, 23 Oct 2008 22:13:43 +0000 (UTC)
Subject: [SciPy-user] Suggestion about algorithm
References: <4900B348.9050304@grinta.net>
Message-ID: <loom.20081023T220624-976@post.gmane.org>


Daniele Nicolodi <daniele <at> grinta.net> writes:

> 
> Hello, i'm going to ask something not strictly related to scipy. Forgive 
> me if this is not appropriate on the mailing list, but i don't know 
> where else i can seek for help, any suggestion is appreciated.
> 
> I'm measuring the quality factor Q of a mechanical oscillator. I use the 
> ring down technique: i excite the oscillator to a big oscillation 
> amplitude so that my read out noise is negligible and then i observe the 
>   decay of the oscillation amplitude during time.

What about Harminv ?

http://ab-initio.mit.edu/wiki/index.php/Harminv

Regards,

  Seb






From stefan at sun.ac.za  Fri Oct 24 03:55:49 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Fri, 24 Oct 2008 09:55:49 +0200
Subject: [SciPy-user] SciPy Sprint Weekend: 1 and 2 November
In-Reply-To: <9457e7c80810240054o3f00cc5fp7528a65af10cc2c7@mail.gmail.com>
References: <9457e7c80810231422m6cab2e75wbcb343472449312@mail.gmail.com>
	<4C4FE880-8830-4153-83CF-7565E5EA4646@cs.toronto.edu>
	<9457e7c80810240054o3f00cc5fp7528a65af10cc2c7@mail.gmail.com>
Message-ID: <9457e7c80810240055m4f53ae56yb4621b5c786f2f22@mail.gmail.com>

Sorry, wrong list.

---------- Forwarded message ----------
From: St?fan van der Walt <stefan at sun.ac.za>
Date: 2008/10/24
Subject: Re: [SciPy-user] SciPy Sprint Weekend: 1 and 2 November
To: SciPy Developers List <scipy-dev at scipy.org>


2008/10/24 David Warde-Farley <dwf at cs.toronto.edu>:
> I think you mean 2008? ;)

Yes, of course!

Event: SciPy Sprint
Date: 1, 2 November 2008
Find us on: irc.freenode.net in channel #scipy

I hope that all of you can join us.  Nathan has put a lot of work into
sparse matrices, David has been working on build systems, other David
improved special functions, we have a brand new spatial module by
Anne, other other David is documenting ndimage, Damian implemented
hierarchical clustering and Tiziano is contributing generalised
eigenproblems.  This is going to be a very useful release -- we must
just get it out the door!

Cheers
St?fan


From stefan at sun.ac.za  Fri Oct 24 03:59:45 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Fri, 24 Oct 2008 09:59:45 +0200
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <gdqrq9$arj$1@ger.gmane.org>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
	<c7009a550810230041s115ad74exa03217d47f2c73ad@mail.gmail.com>
	<gdqrq9$arj$1@ger.gmane.org>
Message-ID: <9457e7c80810240059m1accd206pd3ece9689cec8d67@mail.gmail.com>

Hi Pauli,

2008/10/23 Pauli Virtanen <pav at iki.fi>:
> I started working on Sphinx stuff for Scipy last week, and got this far:
>
>        http://www.iki.fi/pav/tmp/scipy-refguide.tar.gz (source)
>        http://www.iki.fi/pav/tmp/scipy-refguide/

This already looks great!  SciPy is a *big* library, and I think this
reference guide will be extremely useful.

Did you see the post about an improved NumpyExt on the Sphinx mailing
list?  Maybe we can incorporate those changes back into our code.

Cheers
St?fan


From millman at berkeley.edu  Fri Oct 24 05:12:32 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Fri, 24 Oct 2008 02:12:32 -0700
Subject: [SciPy-user] SciPy Sprint Weekend: 1 and 2 November
In-Reply-To: <9457e7c80810231422m6cab2e75wbcb343472449312@mail.gmail.com>
References: <9457e7c80810231422m6cab2e75wbcb343472449312@mail.gmail.com>
Message-ID: <c7009a550810240212p30ff7b40ja43c71caa4a6dfea@mail.gmail.com>

On Thu, Oct 23, 2008 at 2:22 PM, St?fan van der Walt <stefan at sun.ac.za> wrote:
> Event: SciPy Sprint
> Date: 1, 2 November 2008
> Find us on: irc.freenode.net in channel #scipy

This is an absolutely excellent idea!  I will organize a sprint at UC
Berkeley for the 1st and 2nd of November and try to rope in as many
people as I am able.  If we start when you stop (and vice versa), we
may be able to pull of 48 consecutive hours of test fixing, code clean
up, and documentation.  I am looking forward to releasing SciPy 0.7
very soon.

Thanks,

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From daniele at grinta.net  Fri Oct 24 06:32:07 2008
From: daniele at grinta.net (Daniele Nicolodi)
Date: Fri, 24 Oct 2008 12:32:07 +0200
Subject: [SciPy-user] Suggestion about algorithm
In-Reply-To: <3d375d730810231120n97f301cs7b3200904f11a615@mail.gmail.com>
References: <4900B348.9050304@grinta.net>
	<3d375d730810231120n97f301cs7b3200904f11a615@mail.gmail.com>
Message-ID: <4901A427.8010504@grinta.net>

Robert Kern wrote:
> Can you just get the oscillating curve itself rather than extracting
> the peaks? It might be easiest just to fit the decaying oscillator
> function to the curve. Your uncertainly may still be large, but
> probably better than what you currently have.

The point into demodulating the oscillation an taking the derivative of 
the amplitude is that in this way every oscillation becomes an 
independent estimation of the decay constant Beta. I'm convinced that a 
fitting procedure is not statistically correct because each point in my 
time series is strongly correlated with the previous ones (the 
characteristic time of the system is of a couple of years).

In other words an external disturbance acting on the oscillator (think 
to it as a kick) influences only a single data point in my estimation of 
Beta with the derivative. I then average all the independent data 
points. With a fit a kick would instead affect the estimation very badly.

Ciao
-- 
Daniele


From lopmart at gmail.com  Fri Oct 24 09:03:43 2008
From: lopmart at gmail.com (Jose Lopez)
Date: Fri, 24 Oct 2008 06:03:43 -0700
Subject: [SciPy-user] error at sparse
Message-ID: <4eeef9d40810240603i5c0c0897y2fdf15b9d9052682@mail.gmail.com>

hi, i am working with scipy before, and i was used the next line code
'Matrix=sparse.lil_matrix((100,100),float)' and did not have any problem,
but now, the python give me the next error 'NameError: name 'sparse' is not
defined '
somebody knows why?


thanks

pd  i have Python 2.5.2 and SciPy 0.6.0.1
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081024/881cb373/attachment.html>

From robert.kern at gmail.com  Fri Oct 24 09:06:05 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 24 Oct 2008 08:06:05 -0500
Subject: [SciPy-user] ndimage starting points
In-Reply-To: <gdqrq9$arj$1@ger.gmane.org>
References: <FCB82D86-7C94-4CB2-967B-08B5A533D84A@cs.toronto.edu>
	<c7009a550810230041s115ad74exa03217d47f2c73ad@mail.gmail.com>
	<gdqrq9$arj$1@ger.gmane.org>
Message-ID: <3d375d730810240606q53b2c05fx873f923101600fb6@mail.gmail.com>

On Thu, Oct 23, 2008 at 16:54, Pauli Virtanen <pav at iki.fi> wrote:
> Thu, 23 Oct 2008 00:41:51 -0700, Jarrod Millman wrote:
>
>> Since you are offering to help out with this, I would like to see
>> someone do the following:
>>
>> Take the existing numarray.ndimage docs:
>>   http://stsdas.stsci.edu/numarray/numarray-1.5.html/module-
> numarray.ndimage.html
>>
>> and the cookbook stuff:
>>
>>   http://www.scipy.org/SciPyPackages/Ndimage
>>
>> and merge what you can into the docstrings and convert the rest into
>> restructured text and commit it to the scipy trunk.  That way we can
>> start working on generating sphinx documentation for scipy.
>
> I started working on Sphinx stuff for Scipy last week, and got this far:
>
>        http://www.iki.fi/pav/tmp/scipy-refguide.tar.gz (source)
>        http://www.iki.fi/pav/tmp/scipy-refguide/
>
> It's a Sphinx framework similar to the Numpy reference guide we started
> working on this summer. There's in principle also a BZR branch for this
> in Launchpad,
>
>        https://code.launchpad.net/~pauli-virtanen/scipy/scipy-refguide
>
> which you could track using Bazaar (http://bazaar-vcs.org/), but for some
> reason Launchpad decided to dislike me today, so it doesn't work now.
> (BTW, should I put this to Scipy SVN, and if yes, where? What about the
> corresponding Numpy documentation?)

If you would like to, sure. If you want to stick with Bazaar, that's
fine with me, too, but I would like to see official links from
somewhere (like the Doc Marathon wiki page) to what you consider to be
the trunk for each.

If you do want to move it over to SVN, go ahead and make
numpy-refguide/ and scipy-refguide/ directories all the way at the
root as siblings to trunk/. Inside them, you can either make branches/
tags/ trunk/ or you can reuse the top-level branches/ and tags/.
Probably the former.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From robert.kern at gmail.com  Fri Oct 24 09:17:34 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 24 Oct 2008 08:17:34 -0500
Subject: [SciPy-user] error at sparse
In-Reply-To: <4eeef9d40810240603i5c0c0897y2fdf15b9d9052682@mail.gmail.com>
References: <4eeef9d40810240603i5c0c0897y2fdf15b9d9052682@mail.gmail.com>
Message-ID: <3d375d730810240617u232fc119mbf0eddf43cd4486c@mail.gmail.com>

On Fri, Oct 24, 2008 at 08:03, Jose Lopez <lopmart at gmail.com> wrote:
> hi, i am working with scipy before, and i was used the next line code
> 'Matrix=sparse.lil_matrix((100,100),float)' and did not have any problem,
> but now, the python give me the next error 'NameError: name 'sparse' is not
> defined '
> somebody knows why?

Most likely you are missing an import of the sparse package. If not,
can you show us a complete example that fails?

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From lopmart at gmail.com  Fri Oct 24 09:42:25 2008
From: lopmart at gmail.com (Jose Lopez)
Date: Fri, 24 Oct 2008 06:42:25 -0700
Subject: [SciPy-user] error at sparse
In-Reply-To: <3d375d730810240617u232fc119mbf0eddf43cd4486c@mail.gmail.com>
References: <4eeef9d40810240603i5c0c0897y2fdf15b9d9052682@mail.gmail.com>
	<3d375d730810240617u232fc119mbf0eddf43cd4486c@mail.gmail.com>
Message-ID: <4eeef9d40810240642u28450542g37f64e864fa32ffa@mail.gmail.com>

my headers are

from pylab import *
from scipy import *

JL

On Fri, Oct 24, 2008 at 6:17 AM, Robert Kern <robert.kern at gmail.com> wrote:

> On Fri, Oct 24, 2008 at 08:03, Jose Lopez <lopmart at gmail.com> wrote:
> > hi, i am working with scipy before, and i was used the next line code
> > 'Matrix=sparse.lil_matrix((100,100),float)' and did not have any problem,
> > but now, the python give me the next error 'NameError: name 'sparse' is
> not
> > defined '
> > somebody knows why?
>
> Most likely you are missing an import of the sparse package. If not,
> can you show us a complete example that fails?
>
> --
> Robert Kern
>
> "I have come to believe that the whole world is an enigma, a harmless
> enigma that is made terrible by our own mad attempt to interpret it as
> though it had an underlying truth."
>  -- Umberto Eco
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081024/19b6e228/attachment.html>

From robert.kern at gmail.com  Fri Oct 24 10:36:34 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 24 Oct 2008 09:36:34 -0500
Subject: [SciPy-user] error at sparse
In-Reply-To: <4eeef9d40810240642u28450542g37f64e864fa32ffa@mail.gmail.com>
References: <4eeef9d40810240603i5c0c0897y2fdf15b9d9052682@mail.gmail.com>
	<3d375d730810240617u232fc119mbf0eddf43cd4486c@mail.gmail.com>
	<4eeef9d40810240642u28450542g37f64e864fa32ffa@mail.gmail.com>
Message-ID: <3d375d730810240736l25ac95acqf38663a20fed59fc@mail.gmail.com>

On Fri, Oct 24, 2008 at 08:42, Jose Lopez <lopmart at gmail.com> wrote:
> my headers are
>
> from pylab import *
> from scipy import *

Right. The subpackages of scipy do not get imported with "from scipy
import *". You need to do

    from scipy import sparse

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From dlrt2 at ast.cam.ac.uk  Fri Oct 24 10:26:19 2008
From: dlrt2 at ast.cam.ac.uk (David Trethewey)
Date: Fri, 24 Oct 2008 15:26:19 +0100
Subject: [SciPy-user] optimize.leastsq - Value Error: shape mismatch
Message-ID: <4901DB0B.3060704@ast.cam.ac.uk>

I get the following error:

Traceback (most recent call last):
  File "M31FeHfit_total.py", line 86, in <module>
    p1,success  = optimize.leastsq(errfunc, p0, args = (X, Y))
  File "/usr/lib/python2.5/site-packages/scipy/optimize/minpack.py", 
line 264, in leastsq
    m = check_func(func,x0,args,n)[0]
  File "/usr/lib/python2.5/site-packages/scipy/optimize/minpack.py", 
line 11, in check_func
    res = atleast_1d(thefunc(*((x0[:numinputs],)+args)))
  File "M31FeHfit_total.py", line 81, in <lambda>
    errfunc = lambda p, x, y: fitfunc(p,x) - y          # Distance to 
the target function

ValueError: shape mismatch: objects cannot be broadcast to a single shape

when running the following code:

    hista = hist(FeH_sub_range,bins)
     print "hista[1] = ",hista[1]
     print "hista[0] = ",hista[0]
     X = numpy.array(hista[1])
     Y = numpy.array(hista[0])
     print X
#fit gaussian
     fitfunc = lambda p, x: (p[0]**2)*exp(-(x-p[1])**2/(2*p[2]**2))  # 
Target function
     errfunc = lambda p, x, y: fitfunc(p,x) - y          # Distance to 
the target function
     doublegauss = lambda q,x: (q[0]**2)*exp(-(x-q[1])**2/(2*q[2]**2)) + 
(q[3]**2)*exp(-(x-q[4])**2/(2*q[5]**2))
     doublegausserr = lambda q,x,y: doublegauss(q,x) - y
    
     p0 = numpy.array([10.0,-2,0.5])
     p1,success  = optimize.leastsq(errfunc, p0, args = (X, Y))

As I am passing the arguments as numpy arrays I don't see why there is a 
problem. Additionally this problem has arisen since changing version of 
Python.

David Trethewey


From dlrt2 at ast.cam.ac.uk  Fri Oct 24 10:31:52 2008
From: dlrt2 at ast.cam.ac.uk (David Trethewey)
Date: Fri, 24 Oct 2008 15:31:52 +0100
Subject: [SciPy-user] optimize.leastsq - Value Error: shape mismatch
Message-ID: <4901DC58.9050400@ast.cam.ac.uk>


Traceback (most recent call last):
 File "M31FeHfit_total.py", line 86, in <module>
   p1,success  = optimize.leastsq(errfunc, p0, args = (X, Y))
 File "/usr/lib/python2.5/site-packages/scipy/optimize/minpack.py", line 
264, in leastsq
   m = check_func(func,x0,args,n)[0]
 File "/usr/lib/python2.5/site-packages/scipy/optimize/minpack.py", line 
11, in check_func
   res = atleast_1d(thefunc(*((x0[:numinputs],)+args)))
 File "M31FeHfit_total.py", line 81, in <lambda>
   errfunc = lambda p, x, y: fitfunc(p,x) - y          # Distance to the 
target function

ValueError: shape mismatch: objects cannot be broadcast to a single shape

when running the following code:

   hista = hist(FeH_sub_range,bins)
    print "hista[1] = ",hista[1]
    print "hista[0] = ",hista[0]
    X = numpy.array(hista[1])
    Y = numpy.array(hista[0])
    print X
#fit gaussian
    fitfunc = lambda p, x: (p[0]**2)*exp(-(x-p[1])**2/(2*p[2]**2))  # 
Target function
    errfunc = lambda p, x, y: fitfunc(p,x) - y          # Distance to 
the target function
    doublegauss = lambda q,x: (q[0]**2)*exp(-(x-q[1])**2/(2*q[2]**2)) + 
(q[3]**2)*exp(-(x-q[4])**2/(2*q[5]**2))
    doublegausserr = lambda q,x,y: doublegauss(q,x) - y
       p0 = numpy.array([10.0,-2,0.5])
    p1,success  = optimize.leastsq(errfunc, p0, args = (X, Y))

As I am passing the arguments as numpy arrays I don't see why there is a 
problem. Additionally this problem has arisen since changing version of 
Python.

David Trethewey


From jdh2358 at gmail.com  Sat Oct 25 09:50:31 2008
From: jdh2358 at gmail.com (John Hunter)
Date: Sat, 25 Oct 2008 08:50:31 -0500
Subject: [SciPy-user] decimate
Message-ID: <88e473830810250650i9100ef0tf52acc3ec1787b8a@mail.gmail.com>

One of the functions I used to use a lot in matlab was decimate, which
downsamples data after doing a low pass filter to prevent aliasing.

It would be nice to have an analogous simple function in scipy which
does the same -- I found the example below implementation on google.
Does this look like somthing suitable for inclusion in scipy, or is
there an equivalently easier way to do it with existing scipy tools.
The context of this question is that we are teaching python for
scientific computing this weekend to some students in Calremont who
are trying to wean themselves from the matlab teat

http://snippets.dzone.com/posts/show/1209


From josef.pktd at gmail.com  Sat Oct 25 14:03:38 2008
From: josef.pktd at gmail.com (joep)
Date: Sat, 25 Oct 2008 11:03:38 -0700 (PDT)
Subject: [SciPy-user] return type of inverse cdf of discrete distribution?
Message-ID: <3aa03dc4-73ba-4198-bdb6-222f9048c433@u57g2000hsf.googlegroups.com>

What should be the return type of the inverse cdf (and inverse
survival function of a discrete distribution?

The problem is handling of inf for boundary and nans for invalid
input. Options are
* return floating point (double) with inf and nans returned as for the
continuous distribution, or
* return integer and throw an exception if return values are inf or
nans (or restricting to open interval (0,1).

Currently, scipy.stats returns integers (long), but the treatment is
not consistent, e.g. instead of nans, zeros are returned for invalid
input and inf on boundary throws casting error.

I just checked in R:
continuous distribution: inverse cdf returns nans and infs, e.g.
> qnorm(c(0.5,1.0,2.0), 0, 25)
[1]   0 Inf NaN

discrete distribution in VGAM: only accept values in (0,1): e.g.

> qpospois(c(0.5,1.0,2.0), 25)
Error in qpospois(c(0.5, 1, 2), 25) : bad input for argument "p"
> qpospois(1.0, 25)
Error in qpospois(1, 25) : bad input for argument "p"
> qpospois(0.0, 25)
Error in qpospois(0, 25) : bad input for argument "p"
> qpospois(c(0.0000001,0.5,0.999999999), 25)
[1]  4 25 60
>

however in stats package in R: no domain checking, and returns nans
and inf

> aa=qpois(c(0.5,1.0,2.0), 25)
Warning message:
In qpois(p, lambda, lower.tail, log.p) : NaNs produced
> aa
[1]  25 Inf NaN

> typeof(aa)
[1] "double"
> aap=qpospois(c(0.5,0.99), 25)
> typeof(aap)
[1] "double"
> aa1=qpois(0.5, 25)
> typeof(aa1)
[1] "double"

both VGAM and stats in R return double.

Changing the return type in scipy.stats discrete distribution to
double would be a break in API, I don't know if this is relevant or if
anybody cares.

An alternative would be to choose the return type depending on the
presence of nans or infs, but that might not be very reliable for
applications.

Josef


From oliphant at enthought.com  Sat Oct 25 15:42:57 2008
From: oliphant at enthought.com (Travis E. Oliphant)
Date: Sat, 25 Oct 2008 14:42:57 -0500
Subject: [SciPy-user] decimate
In-Reply-To: <88e473830810250650i9100ef0tf52acc3ec1787b8a@mail.gmail.com>
References: <88e473830810250650i9100ef0tf52acc3ec1787b8a@mail.gmail.com>
Message-ID: <490376C1.4050006@enthought.com>

John Hunter wrote:
> One of the functions I used to use a lot in matlab was decimate, which
> downsamples data after doing a low pass filter to prevent aliasing.
>   

You can use the function resample in scipy.signal.  It uses a Fourier 
method for the low-pass filter. 

resample(x, len(x)/q)

should give a similar result as decimate(x, q)

-Travis



From wbaxter at gmail.com  Sat Oct 25 15:53:10 2008
From: wbaxter at gmail.com (Bill Baxter)
Date: Sun, 26 Oct 2008 04:53:10 +0900
Subject: [SciPy-user] decimate
In-Reply-To: <490376C1.4050006@enthought.com>
References: <88e473830810250650i9100ef0tf52acc3ec1787b8a@mail.gmail.com>
	<490376C1.4050006@enthought.com>
Message-ID: <e86a5fd00810251253j4b9b881cxf039d51b354811f0@mail.gmail.com>

On Sun, Oct 26, 2008 at 4:42 AM, Travis E. Oliphant
<oliphant at enthought.com> wrote:
> John Hunter wrote:
>> One of the functions I used to use a lot in matlab was decimate, which
>> downsamples data after doing a low pass filter to prevent aliasing.
>>
>
> You can use the function resample in scipy.signal.  It uses a Fourier
> method for the low-pass filter.
>
> resample(x, len(x)/q)
>
> should give a similar result as decimate(x, q)

Ooh, wiki fodder!
Pasted this tidbit here: http://www.scipy.org/NumPy_for_Matlab_Users

--bb


From nono.231 at gmail.com  Sat Oct 25 16:14:46 2008
From: nono.231 at gmail.com (I. Soumpasis)
Date: Sat, 25 Oct 2008 21:14:46 +0100
Subject: [SciPy-user] ANN: Python programs for epidemic modelling
Message-ID: <3ff92a550810251314m18d0596fxffb0a09658a21260@mail.gmail.com>

Dear lists,

DeductiveThinking.com now provides the Python programs for the book of M.
Keeling & P. Rohani "Modeling Infectious Diseases in Humans and Animals",
Princeton University Press, 2008. The book has on-line material which
includes programs for different models in various programming languages and
mathematical tools such as, "C++, FORTRAN and Matlab, while some are also
coded in the web-based Java programming language to allow readers to quickly
experiment with these types of models", as it is stated at the website. The
Python version of the programs were written long ago and submitted to the
book's on line material website (available soon). The Python programs with
the basic equations modelled and the results in figures were now uploaded on
a special wiki page of DeductiveThinking.com.

Since, the programs are heavily using numpy, scipy and matplotlib libraries,
I send this announcement to all the three lists and the main python-list;
sorry for double-posting. The announcement with the related links is
uploaded here http://blog.deductivethinking.com/?p=29. The programs are at
http://wiki.deductivethinking.com/wiki/Python_Programs_for_Modelling_Infectious_Diseases_book.
For those who are interested on modelling and epidemiology, they can take a
look at the main site (http://deductivethinking.com) or the main page of the
wiki (http://wiki.deductivethinking.com) and follow the epidemiology links.
The website is in its beginning, so limited information is uploaded up to
now.

Thanks for your time and I hope it will be useful for some people,
Best Regards,
Ilias Soumpasis
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081025/2ffb172e/attachment.html>

From haase at msg.ucsf.edu  Sat Oct 25 16:21:50 2008
From: haase at msg.ucsf.edu (Sebastian Haase)
Date: Sat, 25 Oct 2008 22:21:50 +0200
Subject: [SciPy-user] decimate
In-Reply-To: <e86a5fd00810251253j4b9b881cxf039d51b354811f0@mail.gmail.com>
References: <88e473830810250650i9100ef0tf52acc3ec1787b8a@mail.gmail.com>
	<490376C1.4050006@enthought.com>
	<e86a5fd00810251253j4b9b881cxf039d51b354811f0@mail.gmail.com>
Message-ID: <bc657ead0810251321y39fd5139sb7b7f93da271e805@mail.gmail.com>

On Sat, Oct 25, 2008 at 9:53 PM, Bill Baxter <wbaxter at gmail.com> wrote:
> On Sun, Oct 26, 2008 at 4:42 AM, Travis E. Oliphant
> <oliphant at enthought.com> wrote:
>> John Hunter wrote:
>>> One of the functions I used to use a lot in matlab was decimate, which
>>> downsamples data after doing a low pass filter to prevent aliasing.
>>>
>>
>> You can use the function resample in scipy.signal.  It uses a Fourier
>> method for the low-pass filter.
>>
>> resample(x, len(x)/q)
>>
>> should give a similar result as decimate(x, q)
>
> Ooh, wiki fodder!
> Pasted this tidbit here: http://www.scipy.org/NumPy_for_Matlab_Users
Should it read   Sci.signal.resample   !?

-Sebastian Haase


From wbaxter at gmail.com  Sat Oct 25 16:27:13 2008
From: wbaxter at gmail.com (Bill Baxter)
Date: Sun, 26 Oct 2008 05:27:13 +0900
Subject: [SciPy-user] decimate
In-Reply-To: <bc657ead0810251321y39fd5139sb7b7f93da271e805@mail.gmail.com>
References: <88e473830810250650i9100ef0tf52acc3ec1787b8a@mail.gmail.com>
	<490376C1.4050006@enthought.com>
	<e86a5fd00810251253j4b9b881cxf039d51b354811f0@mail.gmail.com>
	<bc657ead0810251321y39fd5139sb7b7f93da271e805@mail.gmail.com>
Message-ID: <e86a5fd00810251327m7fe7a0a0med46f5207cd7d9cb@mail.gmail.com>

On Sun, Oct 26, 2008 at 5:21 AM, Sebastian Haase <haase at msg.ucsf.edu> wrote:
> On Sat, Oct 25, 2008 at 9:53 PM, Bill Baxter <wbaxter at gmail.com> wrote:
>> On Sun, Oct 26, 2008 at 4:42 AM, Travis E. Oliphant
>> <oliphant at enthought.com> wrote:
>>> John Hunter wrote:
>>>> One of the functions I used to use a lot in matlab was decimate, which
>>>> downsamples data after doing a low pass filter to prevent aliasing.
>>>>
>>>
>>> You can use the function resample in scipy.signal.  It uses a Fourier
>>> method for the low-pass filter.
>>>
>>> resample(x, len(x)/q)
>>>
>>> should give a similar result as decimate(x, q)
>>
>> Ooh, wiki fodder!
>> Pasted this tidbit here: http://www.scipy.org/NumPy_for_Matlab_Users
> Should it read   Sci.signal.resample   !?

Oops.  I guess so.  I missed the mention of the package there.

--bb


From aisaac at american.edu  Sat Oct 25 17:46:25 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Sat, 25 Oct 2008 17:46:25 -0400
Subject: [SciPy-user] ANN: Python programs for epidemic modelling
In-Reply-To: <3ff92a550810251314m18d0596fxffb0a09658a21260@mail.gmail.com>
References: <3ff92a550810251314m18d0596fxffb0a09658a21260@mail.gmail.com>
Message-ID: <490393B1.7030904@american.edu>

On 10/25/2008 4:14 PM I. Soumpasis apparently wrote:
> http://blog.deductivethinking.com/?p=29

This is cool.
But I do not see a license.
May I hope this is released under the new BSD license,
like the packages it depends on?

Thanks,
Alan Isaac



From nono.231 at gmail.com  Sat Oct 25 18:07:53 2008
From: nono.231 at gmail.com (I. Soumpasis)
Date: Sat, 25 Oct 2008 23:07:53 +0100
Subject: [SciPy-user] [Numpy-discussion] ANN: Python programs for
	epidemic modelling
In-Reply-To: <490393B1.7030904@american.edu>
References: <3ff92a550810251314m18d0596fxffb0a09658a21260@mail.gmail.com>
	<490393B1.7030904@american.edu>
Message-ID: <3ff92a550810251507q3696aa15x46f5b96684ff7f3c@mail.gmail.com>

2008/10/25 Alan G Isaac <aisaac at american.edu>

> On 10/25/2008 4:14 PM I. Soumpasis apparently wrote:
> > http://blog.deductivethinking.com/?p=29
>
> This is cool.
> But I do not see a license.
> May I hope this is released under the new BSD license,
> like the packages it depends on?
>
> The programs are GPL licensed. More info on the section of copyrights
http://wiki.deductivethinking.com/wiki/Deductive_Thinking:Copyrights.

I hope it is ok,
Ilias
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081025/202067fa/attachment.html>

From aisaac at american.edu  Sat Oct 25 23:55:43 2008
From: aisaac at american.edu (Alan G Isaac)
Date: Sat, 25 Oct 2008 23:55:43 -0400
Subject: [SciPy-user] [Numpy-discussion] ANN: Python programs
 for	epidemic modelling
In-Reply-To: <3ff92a550810251507q3696aa15x46f5b96684ff7f3c@mail.gmail.com>
References: <3ff92a550810251314m18d0596fxffb0a09658a21260@mail.gmail.com>	<490393B1.7030904@american.edu>
	<3ff92a550810251507q3696aa15x46f5b96684ff7f3c@mail.gmail.com>
Message-ID: <4903EA3F.2070504@american.edu>

On 10/25/2008 6:07 PM I. Soumpasis wrote:
> The programs are GPL licensed. More info on the section of copyrights
 > http://wiki.deductivethinking.com/wiki/Deductive_Thinking:Copyrights.
> I hope it is ok,

Well, that depends what you mean by "ok".

Obviously, the author picks the license s/he prefers.
But a GPL license means that some people will avoid
your code, so you make wish to make sure you thought
the licensing issue for this code carefully.

As a point of comparison,
note that all your package dependencies have
a new BSD license.

Alan Isaac



From pwdavenport at gmail.com  Sat Oct 25 08:40:43 2008
From: pwdavenport at gmail.com (Peter Davenport)
Date: Sat, 25 Oct 2008 12:40:43 +0000 (UTC)
Subject: [SciPy-user] array indices, column names
Message-ID: <loom.20081025T121609-681@post.gmane.org>

I want to read in a tab separated data table (from excel), with row and column
names, then remove or add sets of columns to this table as required. 
The key here is to be able to read in a list of columns that I want to remove,
rather than doing it one at a time.

I'm a python novice who's dabbled in tcl and perl, I'm trying python and scipy
since they apparently handle matrixes better. 
I can't find a way of doing this in excel with data filters - hence my attempt
at python and scipy. 

I hope this post is clear, pls tell me if otherwise.

As far as I can see from the scipy cookbook and documentation, my best bet seems
to be as follows:

To read my table into python from a tab separated txt file and set it as an
array.

    A = array([[columnnames, a, b, c, d]
        [row1, 1, 2, 3, 4]
        [row2, 5, 6, 7, 8]
        [row3, 9, 10, 11, 12]])

I can then get specific columns using,
    
    b=A[,0:2]

and presumably it won't be too complex to create new array without these columns
in a manner such as:

    badcolumns = [1:2:4]
    c=A[,badcolumns]

My issues is then, how do I turn a list of column names into a list of column
indices and then remove them from an array?

i.e. starting with a list of column names:

    badcnames = [a,b,d]

how do i get a list of indices:

    badcolumns = [1:2:4]

so that I can then remove them from the array.


Any advice greatly appreciated,

Pete



From tom.denniston at gmail.com  Sun Oct 26 11:05:38 2008
From: tom.denniston at gmail.com (Tom Denniston)
Date: Sun, 26 Oct 2008 10:05:38 -0500
Subject: [SciPy-user] array indices, column names
In-Reply-To: <loom.20081025T121609-681@post.gmane.org>
References: <loom.20081025T121609-681@post.gmane.org>
Message-ID: <E6DC8D5F-2EE1-421F-99F7-C8BD7405CCE0@gmail.com>

There are a number of things to say here:

First, if you want to read directly from .xls files you might want to  
try pyexcelerator.

As I understand it what you are try to do is find what index  
corresponds to the columnames you care about in the larger list of all  
the column names.  If this is the case use numpy.searchsorted.

Finally, you might find it easier to read your info into a recaraay  
using numpy.fromiter and then you will be able to access columns by  
name.

Hope this helps

--Tom


On Oct 25, 2008, at 7:40 AM, Peter Davenport <pwdavenport at gmail.com>  
wrote:

> I want to read in a tab separated data table (from excel), with row  
> and column
> names, then remove or add sets of columns to this table as required.
> The key here is to be able to read in a list of columns that I want  
> to remove,
> rather than doing it one at a time.
>
> I'm a python novice who's dabbled in tcl and perl, I'm trying python  
> and scipy
> since they apparently handle matrixes better.
> I can't find a way of doing this in excel with data filters - hence  
> my attempt
> at python and scipy.
>
> I hope this post is clear, pls tell me if otherwise.
>
> As far as I can see from the scipy cookbook and documentation, my  
> best bet seems
> to be as follows:
>
> To read my table into python from a tab separated txt file and set  
> it as an
> array.
>
>    A = array([[columnnames, a, b, c, d]
>        [row1, 1, 2, 3, 4]
>        [row2, 5, 6, 7, 8]
>        [row3, 9, 10, 11, 12]])
>
> I can then get specific columns using,
>
>    b=A[,0:2]
>
> and presumably it won't be too complex to create new array without  
> these columns
> in a manner such as:
>
>    badcolumns = [1:2:4]
>    c=A[,badcolumns]
>
> My issues is then, how do I turn a list of column names into a list  
> of column
> indices and then remove them from an array?
>
> i.e. starting with a list of column names:
>
>    badcnames = [a,b,d]
>
> how do i get a list of indices:
>
>    badcolumns = [1:2:4]
>
> so that I can then remove them from the array.
>
>
> Any advice greatly appreciated,
>
> Pete
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user


From voodoochild2006 at gmail.com  Sun Oct 26 12:50:16 2008
From: voodoochild2006 at gmail.com (Ritchie)
Date: Sun, 26 Oct 2008 09:50:16 -0700
Subject: [SciPy-user] scipy 0.6 installation problem on os x 10.5.5
Message-ID: <f5e602630810260950n6ee5ea11wd00fbdcabcc40291@mail.gmail.com>

I'm trying to install scipy 0.6 on my leopard MBP, after a
python setup.py build

as root, I got the following error message :

 compiler: gcc -fno-strict-aliasing -Wno-long-double -no-cpp-precomp
-mno-fused-madd -fno-common -dynamic -DNDEBUG -g -Os -Wall
-Wstrict-prototypes -DMACOSX -I/usr/include/ffi -DENABLE_DTRACE -arch i386
-arch ppc -pipe

creating build/temp.macosx-10.5-i386-2.5/scipy/optimize/Zeros
compile options: '-c'
gcc: scipy/optimize/Zeros/brenth.c
cc1: error: unrecognized command line option "-Wno-long-double"
cc1: error: unrecognized command line option "-Wno-long-double"
lipo: can't open input file: /var/tmp//ccaG7ngI.out (No such file or
directory)
cc1: error: unrecognized command line option "-Wno-long-double"
cc1: error: unrecognized command line option "-Wno-long-double"
lipo: can't open input file: /var/tmp//ccaG7ngI.out (No such file or
directory)
error: Command "gcc -fno-strict-aliasing -Wno-long-double -no-cpp-precomp
-mno-fused-madd -fno-common -dynamic -DNDEBUG -g -Os -Wall
-Wstrict-prototypes -DMACOSX -I/usr/include/ffi -DENABLE_DTRACE -arch i386
-arch ppc -pipe -c scipy/optimize/Zeros/brenth.c -o
build/temp.macosx-10.5-i386-2.5/scipy/optimize/Zeros/brenth.o" failed with
exit status 1


I googled for a while, could not find any solution. Anybody has any idea?
I'm using xcode 3.1.1, gcc-4.2.1, gfortran-4.2.3.

Another problem is that if I do a

sudo python setup.py build

as a normal user, I will have the following error message :

  adding 'scipy/sparse/linalg/dsolve/umfpack/umfpack.i' to sources.
swig: scipy/sparse/linalg/dsolve/umfpack/umfpack.i
swig -python -o
build/src.macosx-10.5-i386-2.5/scipy/sparse/linalg/dsolve/umfpack/_umfpack_wrap.c
-outdir build/src.macosx-10.5-i386-2.5/scipy/sparse/linalg/dsolve/umfpack
scipy/sparse/linalg/dsolve/umfpack/umfpack.i
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:192: Error: Unable to find
'umfpack.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:193: Error: Unable to find
'umfpack_solve.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:194: Error: Unable to find
'umfpack_defaults.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:195: Error: Unable to find
'umfpack_triplet_to_col.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:196: Error: Unable to find
'umfpack_col_to_triplet.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:197: Error: Unable to find
'umfpack_transpose.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:198: Error: Unable to find
'umfpack_scale.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:200: Error: Unable to find
'umfpack_report_symbolic.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:201: Error: Unable to find
'umfpack_report_numeric.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:202: Error: Unable to find
'umfpack_report_info.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:203: Error: Unable to find
'umfpack_report_control.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:215: Error: Unable to find
'umfpack_symbolic.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:216: Error: Unable to find
'umfpack_numeric.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:225: Error: Unable to find
'umfpack_free_symbolic.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:226: Error: Unable to find
'umfpack_free_numeric.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:248: Error: Unable to find
'umfpack_get_lunz.h'
scipy/sparse/linalg/dsolve/umfpack/umfpack.i:272: Error: Unable to find
'umfpack_get_numeric.h'
error: command 'swig' failed with exit status 1


even though I updated and export my $UMFPACK variable. However root does not
have this problem.


Thank you
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081026/5be4a40f/attachment.html>

From keflavich at gmail.com  Sun Oct 26 15:33:40 2008
From: keflavich at gmail.com (Keflavich)
Date: Sun, 26 Oct 2008 12:33:40 -0700 (PDT)
Subject: [SciPy-user] Event handling, API programming
Message-ID: <96169678-2695-4e01-8d66-af24cca835bd@w24g2000prd.googlegroups.com>

Hi, I'm trying to make myself a set of widgets for the first time.
I've gotten to the point that I can draw rectangles and lines and make
them do the right things when re-drawing figures, zooming, etc., but
I'm still a little lost on some points, and I haven't found any really
good documentation.

So, first question: Where should I go for documentation first?

I've been using examples, e.g. widgets.py, and the pygtk event
handling page, http://www.pygtk.org/pygtk2tutorial/sec-EventHandling.html.
This page was a useful explanation of the stuff in widgets.py:
http://www.nabble.com/some-API-documentation-td16204232.html.

Second question:  I have two subplots of different data with the same
dimensions.  I'd like to zoom in to the same region on both figures
when I use zoom-to-box on either one.  How can I do this?  (I'm using
tkAgg)

Thanks,
Adam


From robert.kern at gmail.com  Sun Oct 26 15:43:36 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Sun, 26 Oct 2008 14:43:36 -0500
Subject: [SciPy-user] Event handling, API programming
In-Reply-To: <96169678-2695-4e01-8d66-af24cca835bd@w24g2000prd.googlegroups.com>
References: <96169678-2695-4e01-8d66-af24cca835bd@w24g2000prd.googlegroups.com>
Message-ID: <3d375d730810261243y60c898e8q88fa3cbd8dfc5697@mail.gmail.com>

On Sun, Oct 26, 2008 at 14:33, Keflavich <keflavich at gmail.com> wrote:
> Hi, I'm trying to make myself a set of widgets for the first time.
> I've gotten to the point that I can draw rectangles and lines and make
> them do the right things when re-drawing figures, zooming, etc., but
> I'm still a little lost on some points, and I haven't found any really
> good documentation.
>
> So, first question: Where should I go for documentation first?
>
> I've been using examples, e.g. widgets.py, and the pygtk event
> handling page, http://www.pygtk.org/pygtk2tutorial/sec-EventHandling.html.
> This page was a useful explanation of the stuff in widgets.py:
> http://www.nabble.com/some-API-documentation-td16204232.html.
>
> Second question:  I have two subplots of different data with the same
> dimensions.  I'd like to zoom in to the same region on both figures
> when I use zoom-to-box on either one.  How can I do this?  (I'm using
> tkAgg)

You will want to ask on the matplotlib list.

  https://lists.sourceforge.net/lists/listinfo/matplotlib-users

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From robert.kern at gmail.com  Sun Oct 26 15:45:30 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Sun, 26 Oct 2008 14:45:30 -0500
Subject: [SciPy-user] scipy 0.6 installation problem on os x 10.5.5
In-Reply-To: <f5e602630810260950n6ee5ea11wd00fbdcabcc40291@mail.gmail.com>
References: <f5e602630810260950n6ee5ea11wd00fbdcabcc40291@mail.gmail.com>
Message-ID: <3d375d730810261245k2e472ec9t94b1527f14a45f26@mail.gmail.com>

On Sun, Oct 26, 2008 at 11:50, Ritchie <voodoochild2006 at gmail.com> wrote:
> I'm trying to install scipy 0.6 on my leopard MBP, after a
> python setup.py build
> as root, I got the following error message :
>  compiler: gcc -fno-strict-aliasing -Wno-long-double -no-cpp-precomp
> -mno-fused-madd -fno-common -dynamic -DNDEBUG -g -Os -Wall
> -Wstrict-prototypes -DMACOSX -I/usr/include/ffi -DENABLE_DTRACE -arch i386
> -arch ppc -pipe
> creating build/temp.macosx-10.5-i386-2.5/scipy/optimize/Zeros
> compile options: '-c'
> gcc: scipy/optimize/Zeros/brenth.c
> cc1: error: unrecognized command line option "-Wno-long-double"
> cc1: error: unrecognized command line option "-Wno-long-double"
> lipo: can't open input file: /var/tmp//ccaG7ngI.out (No such file or
> directory)
> cc1: error: unrecognized command line option "-Wno-long-double"
> cc1: error: unrecognized command line option "-Wno-long-double"
> lipo: can't open input file: /var/tmp//ccaG7ngI.out (No such file or
> directory)
> error: Command "gcc -fno-strict-aliasing -Wno-long-double -no-cpp-precomp
> -mno-fused-madd -fno-common -dynamic -DNDEBUG -g -Os -Wall
> -Wstrict-prototypes -DMACOSX -I/usr/include/ffi -DENABLE_DTRACE -arch i386
> -arch ppc -pipe -c scipy/optimize/Zeros/brenth.c -o
> build/temp.macosx-10.5-i386-2.5/scipy/optimize/Zeros/brenth.o" failed with
> exit status 1
>
> I googled for a while, could not find any solution. Anybody has any idea?
> I'm using xcode 3.1.1, gcc-4.2.1, gfortran-4.2.3.

That's your problem. You should use the same version of gcc (4.0.1)
that compiled your Python executable. It looks like gcc dropped a flag
in the intervening period, but Python tries to use the same flags it
was built with to build extensions. I haven't had any problem mixing
gcc 4.0.1 with gfortran 4.2.x. At least for building scipy.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From voodoochild2006 at gmail.com  Sun Oct 26 17:44:32 2008
From: voodoochild2006 at gmail.com (Ritchie Cai)
Date: Sun, 26 Oct 2008 14:44:32 -0700
Subject: [SciPy-user] scipy 0.6 installation problem on os x 10.5.5
In-Reply-To: <3d375d730810261245k2e472ec9t94b1527f14a45f26@mail.gmail.com>
References: <f5e602630810260950n6ee5ea11wd00fbdcabcc40291@mail.gmail.com>
	<3d375d730810261245k2e472ec9t94b1527f14a45f26@mail.gmail.com>
Message-ID: <BA2AD880-6AE0-4CB2-9450-64D41DB88AE8@gmail.com>

Thanks, I just got it to work. Actually I tried 4.0 before, but I was  
just doing a simple easy_install scipy, and it gives some other error  
messages. I guess I should have compiled from source like this:

export MACOSX_DEPLOYMENT_TARGET=10.5
     cd ../scipy
     python setup.py build_src build_clib --fcompiler=gnu95 build_ext  
--fcompiler=gnu95 build
     sudo python setup.py install

On Oct 26, 2008, at 12:45 PM, Robert Kern wrote:

> On Sun, Oct 26, 2008 at 11:50, Ritchie <voodoochild2006 at gmail.com>  
> wrote:
>> I'm trying to install scipy 0.6 on my leopard MBP, after a
>> python setup.py build
>> as root, I got the following error message :
>> compiler: gcc -fno-strict-aliasing -Wno-long-double -no-cpp-precomp
>> -mno-fused-madd -fno-common -dynamic -DNDEBUG -g -Os -Wall
>> -Wstrict-prototypes -DMACOSX -I/usr/include/ffi -DENABLE_DTRACE - 
>> arch i386
>> -arch ppc -pipe
>> creating build/temp.macosx-10.5-i386-2.5/scipy/optimize/Zeros
>> compile options: '-c'
>> gcc: scipy/optimize/Zeros/brenth.c
>> cc1: error: unrecognized command line option "-Wno-long-double"
>> cc1: error: unrecognized command line option "-Wno-long-double"
>> lipo: can't open input file: /var/tmp//ccaG7ngI.out (No such file or
>> directory)
>> cc1: error: unrecognized command line option "-Wno-long-double"
>> cc1: error: unrecognized command line option "-Wno-long-double"
>> lipo: can't open input file: /var/tmp//ccaG7ngI.out (No such file or
>> directory)
>> error: Command "gcc -fno-strict-aliasing -Wno-long-double -no-cpp- 
>> precomp
>> -mno-fused-madd -fno-common -dynamic -DNDEBUG -g -Os -Wall
>> -Wstrict-prototypes -DMACOSX -I/usr/include/ffi -DENABLE_DTRACE - 
>> arch i386
>> -arch ppc -pipe -c scipy/optimize/Zeros/brenth.c -o
>> build/temp.macosx-10.5-i386-2.5/scipy/optimize/Zeros/brenth.o"  
>> failed with
>> exit status 1
>>
>> I googled for a while, could not find any solution. Anybody has any  
>> idea?
>> I'm using xcode 3.1.1, gcc-4.2.1, gfortran-4.2.3.
>
> That's your problem. You should use the same version of gcc (4.0.1)
> that compiled your Python executable. It looks like gcc dropped a flag
> in the intervening period, but Python tries to use the same flags it
> was built with to build extensions. I haven't had any problem mixing
> gcc 4.0.1 with gfortran 4.2.x. At least for building scipy.
>
> -- 
> Robert Kern
>
> "I have come to believe that the whole world is an enigma, a harmless
> enigma that is made terrible by our own mad attempt to interpret it as
> though it had an underlying truth."
>  -- Umberto Eco
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081026/4cae99a7/attachment.html>

From markbak at gmail.com  Mon Oct 27 11:00:17 2008
From: markbak at gmail.com (Mark Bakker)
Date: Mon, 27 Oct 2008 16:00:17 +0100
Subject: [SciPy-user] what happened to Numpy-Discussion on Google?
Message-ID: <6946b9500810270800v2679ffe2hde1e057b093db44f@mail.gmail.com>

Does anybody know what happened to the Numpy-Discussion group on Google?

It was quite helpful (and very quick with replies).

Thanks, Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081027/09332b48/attachment.html>

From robert.kern at gmail.com  Mon Oct 27 12:35:08 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 27 Oct 2008 11:35:08 -0500
Subject: [SciPy-user] what happened to Numpy-Discussion on Google?
In-Reply-To: <6946b9500810270800v2679ffe2hde1e057b093db44f@mail.gmail.com>
References: <6946b9500810270800v2679ffe2hde1e057b093db44f@mail.gmail.com>
Message-ID: <3d375d730810270935u1ea1cc84g7f20aec5880a7b6e@mail.gmail.com>

On Mon, Oct 27, 2008 at 10:00, Mark Bakker <markbak at gmail.com> wrote:
> Does anybody know what happened to the Numpy-Discussion group on Google?
>
> It was quite helpful (and very quick with replies).

Apparently a number of groups (possibly just ones which are gatewayed
to other mailing lists like Numpy-Discussion is gatewayed to
numpy-discussion at scipy.org) simply disappeared from Google recently.
You can subscribe to the actual mailing list here:

  http://www.scipy.org/Mailing_Lists

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From rcsqtc at iiqab.csic.es  Mon Oct 27 13:11:11 2008
From: rcsqtc at iiqab.csic.es (Ramon Crehuet)
Date: Mon, 27 Oct 2008 18:11:11 +0100
Subject: [SciPy-user] Problem with numeric array bounds
Message-ID: <4905F62F.8090407@iiqab.csic.es>

Dear all,
I know Numeric is no longer supported, but we are using an application
(nMoldyn) that uses it, and we are encoutering an unusual problem.
I am using the Numeric 23.8.2 version to create arrays, but there is
something wrong when it comes to create a subarray:
----------------------------------------------------------------------
Python 2.5 (r25:51908, Jan 10 2008, 18:01:52)
[GCC 4.1.2 20061115 (prerelease) (SUSE Linux)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import Numeric
>>> a=Numeric.array([7,5,1])
>>> a
array([7, 5, 1])
>>> a[0:]
array([7, 5])
----------------------------------------------------------------------

The subarray ends at the penultimate element when using ':' On
the other hand, there is no problem when defining sublists. Can be
something wrong with python and/or Numeric installation?

Thanks in advance,

Ramon




From robert.kern at gmail.com  Mon Oct 27 13:38:18 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 27 Oct 2008 12:38:18 -0500
Subject: [SciPy-user] Problem with numeric array bounds
In-Reply-To: <4905F62F.8090407@iiqab.csic.es>
References: <4905F62F.8090407@iiqab.csic.es>
Message-ID: <3d375d730810271038o6a96d774ibd7d369ea7e26915@mail.gmail.com>

On Mon, Oct 27, 2008 at 12:11, Ramon Crehuet <rcsqtc at iiqab.csic.es> wrote:
> Dear all,
> I know Numeric is no longer supported, but we are using an application
> (nMoldyn) that uses it, and we are encoutering an unusual problem.
> I am using the Numeric 23.8.2 version to create arrays, but there is
> something wrong when it comes to create a subarray:
> ----------------------------------------------------------------------
> Python 2.5 (r25:51908, Jan 10 2008, 18:01:52)
> [GCC 4.1.2 20061115 (prerelease) (SUSE Linux)] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
>>>> import Numeric
>>>> a=Numeric.array([7,5,1])
>>>> a
> array([7, 5, 1])
>>>> a[0:]
> array([7, 5])
> ----------------------------------------------------------------------
>
> The subarray ends at the penultimate element when using ':' On
> the other hand, there is no problem when defining sublists. Can be
> something wrong with python and/or Numeric installation?

Try Numeric 24.2. On my machine, it gives the correct results:

>>> import Numeric
>>> a = Numeric.array([7,5,1])
>>> a[0:]
array([7, 5, 1])

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From robert.kern at gmail.com  Mon Oct 27 16:14:59 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 27 Oct 2008 15:14:59 -0500
Subject: [SciPy-user] Some help with chisquare
In-Reply-To: <3d381e170810221055q262d308ah1377a0f0f0fc6c7d@mail.gmail.com>
References: <3d381e170810221055q262d308ah1377a0f0f0fc6c7d@mail.gmail.com>
Message-ID: <3d375d730810271314v271bf2edna316c54025c2e395@mail.gmail.com>

On Wed, Oct 22, 2008 at 12:55, Erik Wickstrom <erik at erikwickstrom.com> wrote:
> Hi,
>
> I'm trying to port an application to python, and want to use scipy to handle
> the statistics.
>
> The app takes several tests and uses chi-square to determines which has the
> highest success rate with a confidence of 95% or better (critical
> values/degrees of freedom).
>
> For example:
> Test a:
> Total trials = 100
> Total successes = 60
>
> Test b:
> Total trials = 105
> Total successes = 46
>
> Test c:
> Total trials = 98
> Total successes = 52
>
> It then puts the data through some sort of chi-square formula (or so the
> comments say) and produces a chi-square value that can be compared against
> the critical values for 95% confidence.
>
> Trouble is, I'm not sure which of the many scipy chi-square functions to
> use, and what data I need to feed into them....

scipy.stats.chisquare() is probably what you want. Pass it arrays of
the actual and expected frequencies for each Test. It will return to
you a Chi^2 value and the associated p-value. If the p-value is <
0.05, then the Chi^2 value is greater than the critical value for the
95% confidence region.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From bsouthey at gmail.com  Mon Oct 27 17:12:21 2008
From: bsouthey at gmail.com (Bruce Southey)
Date: Mon, 27 Oct 2008 16:12:21 -0500
Subject: [SciPy-user] Some help with chisquare
In-Reply-To: <3d375d730810271314v271bf2edna316c54025c2e395@mail.gmail.com>
References: <3d381e170810221055q262d308ah1377a0f0f0fc6c7d@mail.gmail.com>
	<3d375d730810271314v271bf2edna316c54025c2e395@mail.gmail.com>
Message-ID: <49062EB5.30004@gmail.com>

Robert Kern wrote:
> On Wed, Oct 22, 2008 at 12:55, Erik Wickstrom <erik at erikwickstrom.com> wrote:
>   
>> Hi,
>>
>> I'm trying to port an application to python, and want to use scipy to handle
>> the statistics.
>>
>> The app takes several tests and uses chi-square to determines which has the
>> highest success rate with a confidence of 95% or better (critical
>> values/degrees of freedom).
>>
>> For example:
>> Test a:
>> Total trials = 100
>> Total successes = 60
>>
>> Test b:
>> Total trials = 105
>> Total successes = 46
>>
>> Test c:
>> Total trials = 98
>> Total successes = 52
>>
>> It then puts the data through some sort of chi-square formula (or so the
>> comments say) and produces a chi-square value that can be compared against
>> the critical values for 95% confidence.
>>
>> Trouble is, I'm not sure which of the many scipy chi-square functions to
>> use, and what data I need to feed into them....
>>     
>
> scipy.stats.chisquare() is probably what you want. Pass it arrays of
> the actual and expected frequencies for each Test. It will return to
> you a Chi^2 value and the associated p-value. If the p-value is <
> 0.05, then the Chi^2 value is greater than the critical value for the
> 95% confidence region.
>
>   

I think there is insufficient information here because the description 
is rather unclear. I think this sounds like the Cochran-Mantel-Haenszel 
test (http://en.wikipedia.org/wiki/Cochran_test). A formula to calculate 
the chi-square value and degrees of freedom would be clearer as well as 
the actual value and p-value returned for the above example.

Bruce




From 302302 at centrum.cz  Mon Oct 27 21:06:04 2008
From: 302302 at centrum.cz (302302)
Date: Tue, 28 Oct 2008 02:06:04 +0100
Subject: [SciPy-user] Matplotlib redraw axis
In-Reply-To: <200810280139.2758@centrum.cz>
References: <200810280132.21533@centrum.cz> <200810280133.21644@centrum.cz>
	<200810280134.9435@centrum.cz> <200810280135.21986@centrum.cz>
	<200810280136.19448@centrum.cz> <200810280137.19596@centrum.cz>
	<200810280138.19689@centrum.cz> <200810280139.2758@centrum.cz>
Message-ID: <200810280206.2166@centrum.cz>

Hi,
I'm dealing with problem how to redraw just label ticks in one certain subplot with in matplotlib. If I change description of axis in the subplot (by .set_yticks() and .set_yticklabels()) I have to redraw whole figure (figure.canvas.draw()) to see the change.
But I need to redraw either just the one subplot with axis description or just the descriptions. 
Is it possible to use there something like "blit" technique?

Thanks for any advice.
Czenek








From jdh2358 at gmail.com  Mon Oct 27 21:14:12 2008
From: jdh2358 at gmail.com (John Hunter)
Date: Mon, 27 Oct 2008 20:14:12 -0500
Subject: [SciPy-user] Matplotlib redraw axis
In-Reply-To: <200810280206.2166@centrum.cz>
References: <200810280132.21533@centrum.cz> <200810280133.21644@centrum.cz>
	<200810280134.9435@centrum.cz> <200810280135.21986@centrum.cz>
	<200810280136.19448@centrum.cz> <200810280137.19596@centrum.cz>
	<200810280138.19689@centrum.cz> <200810280139.2758@centrum.cz>
	<200810280206.2166@centrum.cz>
Message-ID: <88e473830810271814x3b3ad789sb9a43e9733ce6cea@mail.gmail.com>

On Mon, Oct 27, 2008 at 8:06 PM, 302302 <302302 at centrum.cz> wrote:
> Hi,
> I'm dealing with problem how to redraw just label ticks in one certain subplot with in matplotlib. If I change description of axis in the subplot (by .set_yticks() and .set_yticklabels()) I have to redraw whole figure (figure.canvas.draw()) to see the change.
> But I need to redraw either just the one subplot with axis description or just the descriptions.
> Is it possible to use there something like "blit" technique?

matplotlib questions should be addressed to the matplotlib-users list at:

  https://lists.sourceforge.net/lists/listinfo/matplotlib-users

we'll be happy to answer your questions over there.

JDH


From 302302 at centrum.cz  Mon Oct 27 21:34:41 2008
From: 302302 at centrum.cz (302302)
Date: Tue, 28 Oct 2008 02:34:41 +0100
Subject: [SciPy-user] Matplotlib redraw axis
In-Reply-To: <88e473830810271814x3b3ad789sb9a43e9733ce6cea@mail.gmail.com>
References: <200810280132.21533@centrum.cz> <200810280133.21644@centrum.cz>
	<200810280206.2166@centrum.cz>
	<88e473830810271814x3b3ad789sb9a43e9733ce6cea@mail.gmail.com>
Message-ID: <200810280234.1338@centrum.cz>

Ok, thanks you for directing me to more appropriate forum.

Czenek





______________________________________________________________
> Od: jdh2358 at gmail.com
> Komu: &quot;SciPy Users List&quot; &lt;scipy-user at scipy.org&gt;
> Datum: 28.10.2008 02:14
> P?edm?t: Re: [SciPy-user] Matplotlib redraw axis
>
>On Mon, Oct 27, 2008 at 8:06 PM, 302302 <302302 at centrum.cz> wrote:
>> Hi,
>> I'm dealing with problem how to redraw just label ticks in one certain subplot with in matplotlib. If I change description of axis in the subplot (by .set_yticks() and .set_yticklabels()) I have to redraw whole figure (figure.canvas.draw()) to see the change.
>> But I need to redraw either just the one subplot with axis description or just the descriptions.
>> Is it possible to use there something like "blit" technique?
>
>matplotlib questions should be addressed to the matplotlib-users list at:
>
>  https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>we'll be happy to answer your questions over there.
>
>JDH
>_______________________________________________
>SciPy-user mailing list
>SciPy-user at scipy.org
>http://projects.scipy.org/mailman/listinfo/scipy-user
>



From mwojc at p.lodz.pl  Tue Oct 28 06:38:32 2008
From: mwojc at p.lodz.pl (Marek Wojciechowski)
Date: Tue, 28 Oct 2008 11:38:32 +0100
Subject: [SciPy-user] combinatorics - all set partitions
Message-ID: <ge6puu$tak$1@ger.gmane.org>

Hi!
I'm trying to find an algorithm (and possibly the python code) implementing 
the problem of finding all possible partitions of the set. 

Example of all partitions for the set { 1, 2, 3 } is:
{ {1}, {2}, {3} }
{ {1, 2}, {3} }
{ {1, 3}, {2} }
{ {1}, {2, 3} }
{ {1, 2, 3} }
but i need general partitioning tool.

I thought maybe someone from the group knows the solution...

Greetings,
-- 
Marek Wojciechowski



From wesmckinn at gmail.com  Tue Oct 28 09:46:24 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Tue, 28 Oct 2008 09:46:24 -0400
Subject: [SciPy-user] combinatorics - all set partitions
In-Reply-To: <ge6puu$tak$1@ger.gmane.org>
References: <ge6puu$tak$1@ger.gmane.org>
Message-ID: <6c476c8a0810280646o694b561bgbf54115de4d23b49@mail.gmail.com>

Have you tried looking in Sage (www.sagemath.org)? I think it has everything
you need:

http://www.sagemath.org/doc/ref/module-sage.combinat.set-partition.html

On Tue, Oct 28, 2008 at 6:38 AM, Marek Wojciechowski <mwojc at p.lodz.pl>wrote:

> Hi!
> I'm trying to find an algorithm (and possibly the python code) implementing
> the problem of finding all possible partitions of the set.
>
> Example of all partitions for the set { 1, 2, 3 } is:
> { {1}, {2}, {3} }
> { {1, 2}, {3} }
> { {1, 3}, {2} }
> { {1}, {2, 3} }
> { {1, 2, 3} }
> but i need general partitioning tool.
>
> I thought maybe someone from the group knows the solution...
>
> Greetings,
> --
> Marek Wojciechowski
>
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081028/045360ba/attachment.html>

From reckoner at gmail.com  Tue Oct 28 12:37:35 2008
From: reckoner at gmail.com (Reckoner)
Date: Tue, 28 Oct 2008 12:37:35 -0400
Subject: [SciPy-user] combinatorics - all set partitions
In-Reply-To: <ge6puu$tak$1@ger.gmane.org>
References: <ge6puu$tak$1@ger.gmane.org>
Message-ID: <a1b53c580810280937n75766320scdd73b3019945095@mail.gmail.com>

see chooser.py

http://code.activestate.com/recipes/302478/
http://mail.python.org/pipermail/python-list/2006-May/383412.html

and more at

http://www.diigo.com/user/reckoner/combinatorics?tab=250



On 10/28/08, Marek Wojciechowski <mwojc at p.lodz.pl> wrote:
> Hi!
>  I'm trying to find an algorithm (and possibly the python code) implementing
>  the problem of finding all possible partitions of the set.
>
>  Example of all partitions for the set { 1, 2, 3 } is:
>  { {1}, {2}, {3} }
>  { {1, 2}, {3} }
>  { {1, 3}, {2} }
>  { {1}, {2, 3} }
>  { {1, 2, 3} }
>  but i need general partitioning tool.
>
>  I thought maybe someone from the group knows the solution...
>
>  Greetings,
>
> --
>  Marek Wojciechowski
>
>  _______________________________________________
>  SciPy-user mailing list
>  SciPy-user at scipy.org
>  http://projects.scipy.org/mailman/listinfo/scipy-user
>


From rcsqtc at iiqab.csic.es  Tue Oct 28 13:15:19 2008
From: rcsqtc at iiqab.csic.es (Ramon Crehuet)
Date: Tue, 28 Oct 2008 18:15:19 +0100
Subject: [SciPy-user] Problem with numeric array bounds
In-Reply-To: <mailman.11.1225213205.2640.scipy-user@scipy.org>
References: <mailman.11.1225213205.2640.scipy-user@scipy.org>
Message-ID: <490748A7.9080303@iiqab.csic.es>

Hi,
Unfortunately I have to stick to Numeric 23.8 because the application is
not fully compatible with Numeric 24. In fact, we have another machine
where Numeric 23.8 is working flowlessly so it must be a compilation
problem. However we have updated gcc (to 4.1.3) but the problem remains.
Any suggestion?
Cheers,
Ramon


> Message: 2
> Date: Mon, 27 Oct 2008 12:38:18 -0500
> From: "Robert Kern" <robert.kern at gmail.com>
> Subject: Re: [SciPy-user] Problem with numeric array bounds
> To: "SciPy Users List" <scipy-user at scipy.org>
> Message-ID:
> 	<3d375d730810271038o6a96d774ibd7d369ea7e26915 at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
> 
> On Mon, Oct 27, 2008 at 12:11, Ramon Crehuet <rcsqtc at iiqab.csic.es> wrote:
>> Dear all,
>> I know Numeric is no longer supported, but we are using an application
>> (nMoldyn) that uses it, and we are encoutering an unusual problem.
>> I am using the Numeric 23.8.2 version to create arrays, but there is
>> something wrong when it comes to create a subarray:
>> ----------------------------------------------------------------------
>> Python 2.5 (r25:51908, Jan 10 2008, 18:01:52)
>> [GCC 4.1.2 20061115 (prerelease) (SUSE Linux)] on linux2
>> Type "help", "copyright", "credits" or "license" for more information.
>>>>> import Numeric
>>>>> a=Numeric.array([7,5,1])
>>>>> a
>> array([7, 5, 1])
>>>>> a[0:]
>> array([7, 5])
>> ----------------------------------------------------------------------
>>
>> The subarray ends at the penultimate element when using ':' On
>> the other hand, there is no problem when defining sublists. Can be
>> something wrong with python and/or Numeric installation?
> 
> Try Numeric 24.2. On my machine, it gives the correct results:
> 
>>>> import Numeric
>>>> a = Numeric.array([7,5,1])
>>>> a[0:]
> array([7, 5, 1])
> 



From robert.kern at gmail.com  Tue Oct 28 13:58:44 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 28 Oct 2008 12:58:44 -0500
Subject: [SciPy-user] Problem with numeric array bounds
In-Reply-To: <490748A7.9080303@iiqab.csic.es>
References: <mailman.11.1225213205.2640.scipy-user@scipy.org>
	<490748A7.9080303@iiqab.csic.es>
Message-ID: <3d375d730810281058q46edd660yf8b06e61eb60712b@mail.gmail.com>

On Tue, Oct 28, 2008 at 12:15, Ramon Crehuet <rcsqtc at iiqab.csic.es> wrote:
> Hi,
> Unfortunately I have to stick to Numeric 23.8 because the application is
> not fully compatible with Numeric 24. In fact, we have another machine
> where Numeric 23.8 is working flowlessly so it must be a compilation
> problem. However we have updated gcc (to 4.1.3) but the problem remains.
> Any suggestion?

Huh. Not really. What kind of machines does it fail and work on, respectively?

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From mwojc at p.lodz.pl  Tue Oct 28 15:11:22 2008
From: mwojc at p.lodz.pl (Marek Wojciechowski)
Date: Tue, 28 Oct 2008 20:11:22 +0100
Subject: [SciPy-user] combinatorics - all set partitions
References: <ge6puu$tak$1@ger.gmane.org>
Message-ID: <ge7o0e$h1u$1@ger.gmane.org>

Marek Wojciechowski wrote:

> Hi!
> I'm trying to find an algorithm (and possibly the python code)
> implementing the problem of finding all possible partitions of the set.
> 
> Example of all partitions for the set { 1, 2, 3 } is:
> { {1}, {2}, {3} }
> { {1, 2}, {3} }
> { {1, 3}, {2} }
> { {1}, {2, 3} }
> { {1, 2, 3} }
> but i need general partitioning tool.
> 
> I thought maybe someone from the group knows the solution...
> 
> Greetings,

Thanks for the answers.. However I've ended up finally with writing the
generator by myself:

from copy import deepcopy
def addelement(partlist, e):
    newpartlist = []
    for part in partlist:
        npart = part + [[e]]
        newpartlist += [npart]
        for i in xrange(len(part)):
            npart = deepcopy(part)
            npart[i] += [e]
            newpartlist += [npart]
    return newpartlist

def partition(n):
    if n == 0: return []
    partlist = [[[1]]]
    for i in xrange(2, n+1):
        partlist = addelement(partlist, i)
    return partlist

print partition(4) 

This seems to give good results and is enough for me. 

Thanks again!
-- 
Marek Wojciechowski



From argriffi at ncsu.edu  Tue Oct 28 15:57:40 2008
From: argriffi at ncsu.edu (alex)
Date: Tue, 28 Oct 2008 15:57:40 -0400
Subject: [SciPy-user] combinatorics - all set partitions
In-Reply-To: <ge7o0e$h1u$1@ger.gmane.org>
References: <ge6puu$tak$1@ger.gmane.org> <ge7o0e$h1u$1@ger.gmane.org>
Message-ID: <49076EB4.2020309@ncsu.edu>

Marek Wojciechowski wrote:
> Marek Wojciechowski wrote:
>
>   
>> Hi!
>> I'm trying to find an algorithm (and possibly the python code)
>> implementing the problem of finding all possible partitions of the set.
>>
>> Example of all partitions for the set { 1, 2, 3 } is:
>> { {1}, {2}, {3} }
>> { {1, 2}, {3} }
>> { {1, 3}, {2} }
>> { {1}, {2, 3} }
>> { {1, 2, 3} }
>> but i need general partitioning tool.
>>
>> I thought maybe someone from the group knows the solution...
>>
>> Greetings,
>>     
>
> Thanks for the answers.. However I've ended up finally with writing the
> generator by myself:
>
> from copy import deepcopy
> def addelement(partlist, e):
>     newpartlist = []
>     for part in partlist:
>         npart = part + [[e]]
>         newpartlist += [npart]
>         for i in xrange(len(part)):
>             npart = deepcopy(part)
>             npart[i] += [e]
>             newpartlist += [npart]
>     return newpartlist
>
> def partition(n):
>     if n == 0: return []
>     partlist = [[[1]]]
>     for i in xrange(2, n+1):
>         partlist = addelement(partlist, i)
>     return partlist
>
> print partition(4) 
>
> This seems to give good results and is enough for me. 
>
> Thanks again!
>   
You might want to call it something other than partition to avoid 
conflicts with the builtin function.


From jmiller at stsci.edu  Tue Oct 28 16:15:22 2008
From: jmiller at stsci.edu (Todd Miller)
Date: Tue, 28 Oct 2008 16:15:22 -0400
Subject: [SciPy-user] ANN:  Cyrano v0.1 demo tool
Message-ID: <490772DA.7000807@stsci.edu>

Cyrano is a simple Tk GUI that helps a speaker interactively demonstrate 
Python code.   Cyrano steps through a demonstration script by "typing" 
it into a shell window while you focus on talking to the audience.   
Cyrano's shell window is based on IPython so a speaker can interject 
spontaneous commands at any time during his demo.   Cyrano supports 
customization of fonts and colors using an RC file.

Cyrano is open source software with a BSD-style license.

The first public release of Cyrano,  v0.1,  is downloadable here:  
http://stsdas.stsci.edu/cyrano

Send questions or comments to jmiller at stsci.edu.

Cheers,
Todd Miller





From xavier.gnata at gmail.com  Wed Oct 29 08:07:03 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Wed, 29 Oct 2008 13:07:03 +0100
Subject: [SciPy-user] run=2339 errors=0 failures=3 on ibex
Message-ID: <490851E7.1020706@gmail.com>

Hi,

I have compiled scipy svn on ubuntu ibex.
It looks fine but I have 3 small failures inn scipy.test() :

======================================================================          

FAIL: 
test_lapack.test_all_lapack                                              
----------------------------------------------------------------------          

Traceback (most recent call 
last):                                             
  File "/var/lib/python-support/python2.5/nose/case.py", line 182, in 
runTest  
    
self.test(*self.arg)                                                       
  File 
"/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py", 
line 41, in 
check_syevr                                                         
    
assert_array_almost_equal(w,exact_w)                                       
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line 
310, in 
assert_array_almost_equal                                                        

    header='Arrays are not almost 
equal')                                      
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line 
295, in 
assert_array_compare                                                             

    raise 
AssertionError(msg)                                                  
AssertionError:                                                                 

Arrays are not almost 
equal                                                    

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769462,  9.18222713], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])              

======================================================================
FAIL: test_lapack.test_all_lapack                                    
----------------------------------------------------------------------
Traceback (most recent call last):                                   
  File "/var/lib/python-support/python2.5/nose/case.py", line 182, in 
runTest
    
self.test(*self.arg)                                                    
  File 
"/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py", 
line 66, in 
check_syevr_irange                                                  
    
assert_array_almost_equal(w,exact_w[rslice])                               
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line 
310, in 
assert_array_almost_equal                                                        

    header='Arrays are not almost 
equal')                                      
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line 
295, in 
assert_array_compare                                                             

    raise 
AssertionError(msg)                                                  
AssertionError:                                                                 

Arrays are not almost 
equal                                                    

(mismatch 33.3333333333%)
 x: array([-0.66992444,  0.48769462,  9.18222713], dtype=float32)
 y: array([-0.66992434,  0.48769389,  9.18223045])

======================================================================
FAIL: test_pbdv (test_basic.TestCephes)
----------------------------------------------------------------------
Traceback (most recent call last):
  File 
"/usr/lib/python2.5/site-packages/scipy/special/tests/test_basic.py", 
line 368, in test_pbdv
    assert_equal(cephes.pbdv(1,0),(0.0,0.0))
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line 
176, in assert_equal
    assert_equal(actual[k], desired[k], 'item=%r\n%s' % (k,err_msg), 
verbose)
  File "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", line 
183, in assert_equal
    raise AssertionError(msg)
AssertionError:
Items are not equal:
item=1

 ACTUAL: 1.0
 DESIRED: 0.0

I haven't started to have a deep look into this test code so far.
Any ideas?

Xavier


From nwagner at iam.uni-stuttgart.de  Wed Oct 29 08:16:30 2008
From: nwagner at iam.uni-stuttgart.de (Nils Wagner)
Date: Wed, 29 Oct 2008 13:16:30 +0100
Subject: [SciPy-user] run=2339 errors=0 failures=3 on ibex
In-Reply-To: <490851E7.1020706@gmail.com>
References: <490851E7.1020706@gmail.com>
Message-ID: <web-113448791@uni-stuttgart.de>

On Wed, 29 Oct 2008 13:07:03 +0100
  Xavier Gnata <xavier.gnata at gmail.com> wrote:
> Hi,
> 
> I have compiled scipy svn on ubuntu ibex.
> It looks fine but I have 3 small failures inn 
>scipy.test() :
> 
> ====================================================================== 
>         
> 
>FAIL: 
> test_lapack.test_all_lapack 
>                                             
> ---------------------------------------------------------------------- 
>         
> 
> Traceback (most recent call 
> last):                                             
>  File "/var/lib/python-support/python2.5/nose/case.py", 
>line 182, in 
> runTest  
>    
> self.test(*self.arg) 
>                                                      
>  File 
> "/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py", 
> line 41, in 
> check_syevr 
>                                                        
>    
> assert_array_almost_equal(w,exact_w) 
>                                      
>  File 
>"/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>line 
> 310, in 
> assert_array_almost_equal 
>                                                       
> 
>    header='Arrays are not almost 
> equal')                                      
>  File 
>"/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>line 
> 295, in 
> assert_array_compare 
>                                                           
> 
> 
>    raise 
> AssertionError(msg) 
>                                                 
> AssertionError: 
>                                                           
>     
> 
> Arrays are not almost 
> equal 
>                                                   
> 
> (mismatch 33.3333333333%)
> x: array([-0.66992444,  0.48769462,  9.18222713], 
>dtype=float32)
> y: array([-0.66992434,  0.48769389,  9.18223045]) 
>             
> 
> ======================================================================
>FAIL: test_lapack.test_all_lapack 
>                                   
> ----------------------------------------------------------------------
> Traceback (most recent call last): 
>                                  
>  File "/var/lib/python-support/python2.5/nose/case.py", 
>line 182, in 
> runTest
>    
> self.test(*self.arg) 
>                                                   
>  File 
> "/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py", 
> line 66, in 
> check_syevr_irange 
>                                                 
>    
> assert_array_almost_equal(w,exact_w[rslice]) 
>                              
>  File 
>"/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>line 
> 310, in 
> assert_array_almost_equal 
>                                                       
> 
>    header='Arrays are not almost 
> equal')                                      
>  File 
>"/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>line 
> 295, in 
> assert_array_compare 
>                                                           
> 
> 
>    raise 
> AssertionError(msg) 
>                                                 
> AssertionError: 
>                                                           
>     
> 
> Arrays are not almost 
> equal 
>                                                   
> 
> (mismatch 33.3333333333%)
> x: array([-0.66992444,  0.48769462,  9.18222713], 
>dtype=float32)
> y: array([-0.66992434,  0.48769389,  9.18223045])
> 
> 

This is a known failure

http://projects.scipy.org/scipy/scipy/ticket/375

However I cannot reproduce the next failure (test_pbdv)

Nils

======================================================================
>FAIL: test_pbdv (test_basic.TestCephes)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>  File 
> "/usr/lib/python2.5/site-packages/scipy/special/tests/test_basic.py", 
> line 368, in test_pbdv
>    assert_equal(cephes.pbdv(1,0),(0.0,0.0))
>  File 
>"/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>line 
> 176, in assert_equal
>    assert_equal(actual[k], desired[k], 'item=%r\n%s' % 
>(k,err_msg), 
> verbose)
>  File 
>"/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>line 
> 183, in assert_equal
>    raise AssertionError(msg)
> AssertionError:
> Items are not equal:
> item=1
> 
> ACTUAL: 1.0
> DESIRED: 0.0
> 
> I haven't started to have a deep look into this test 
>code so far.
> Any ideas?
> 
> Xavier
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
  


From millman at berkeley.edu  Wed Oct 29 10:53:11 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Wed, 29 Oct 2008 07:53:11 -0700
Subject: [SciPy-user] ANN: NumPy 1.2.1
In-Reply-To: <c7009a550810290752p7bb1a282qe682db2efd7e262c@mail.gmail.com>
References: <c7009a550810290752p7bb1a282qe682db2efd7e262c@mail.gmail.com>
Message-ID: <c7009a550810290753j6e92e573nca7306b35458aade@mail.gmail.com>

I'm pleased to announce the release of NumPy 1.2.1.

NumPy is the fundamental package needed for scientific computing with
Python.  It contains:

 * a powerful N-dimensional array object
 * sophisticated (broadcasting) functions
 * basic linear algebra functions
 * basic Fourier transforms
 * sophisticated random number capabilities
 * tools for integrating Fortran code.

Besides it's obvious scientific uses, NumPy can also be used as an
efficient multi-dimensional container of generic data. Arbitrary
data-types can be defined. This allows NumPy to seamlessly and
speedily integrate with a wide-variety of databases.

This bugfix release comes almost one month after the 1.2.0 release.
Please note that NumPy 1.2.1 requires Python 2.4 or greater.

For information, please see the release notes:
https://sourceforge.net/project/shownotes.php?release_id=636728&group_id=1369

You can download the release from here:
https://sourceforge.net/project/showfiles.php?group_id=1369

Thank you to everybody who contributed to this release.

Enjoy,

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From xavier.gnata at gmail.com  Wed Oct 29 12:49:35 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Wed, 29 Oct 2008 17:49:35 +0100
Subject: [SciPy-user] run=2339 errors=0 failures=3 on ibex
In-Reply-To: <web-113448791@uni-stuttgart.de>
References: <490851E7.1020706@gmail.com> <web-113448791@uni-stuttgart.de>
Message-ID: <4908941F.6000901@gmail.com>

Ok so there is a simple fix which is fully correct:

-assert_array_almost_equal(x, y, decimal=6, err_msg='', verbose=True)

+assert_array_almost_equal(x, y, decimal=5, err_msg='', verbose=True)

This fix has been proposed a long time ago (07/04/07 )
Where is the problem preventing to merge it?

Nils is not able to reproducec the last faillure. I have to try to understand why...


Xavier

>> Hi,
>>
>> I have compiled scipy svn on ubuntu ibex.
>> It looks fine but I have 3 small failures inn 
>> scipy.test() :
>>
>> ====================================================================== 
>>         
>>
>> FAIL: 
>> test_lapack.test_all_lapack 
>>                                             
>> ---------------------------------------------------------------------- 
>>         
>>
>> Traceback (most recent call 
>> last):                                             
>>  File "/var/lib/python-support/python2.5/nose/case.py", 
>> line 182, in 
>> runTest  
>>    
>> self.test(*self.arg) 
>>                                                      
>>  File 
>> "/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py", 
>> line 41, in 
>> check_syevr 
>>                                                        
>>    
>> assert_array_almost_equal(w,exact_w) 
>>                                      
>>  File 
>> "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>> line 
>> 310, in 
>> assert_array_almost_equal 
>>                                                       
>>
>>    header='Arrays are not almost 
>> equal')                                      
>>  File 
>> "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>> line 
>> 295, in 
>> assert_array_compare 
>>                                                           
>>
>>
>>    raise 
>> AssertionError(msg) 
>>                                                 
>> AssertionError: 
>>                                                           
>>     
>>
>> Arrays are not almost 
>> equal 
>>                                                   
>>
>> (mismatch 33.3333333333%)
>> x: array([-0.66992444,  0.48769462,  9.18222713], 
>> dtype=float32)
>> y: array([-0.66992434,  0.48769389,  9.18223045]) 
>>             
>>
>> ======================================================================
>> FAIL: test_lapack.test_all_lapack 
>>                                   
>> ----------------------------------------------------------------------
>> Traceback (most recent call last): 
>>                                  
>>  File "/var/lib/python-support/python2.5/nose/case.py", 
>> line 182, in 
>> runTest
>>    
>> self.test(*self.arg) 
>>                                                   
>>  File 
>> "/usr/lib/python2.5/site-packages/scipy/lib/lapack/tests/esv_tests.py", 
>> line 66, in 
>> check_syevr_irange 
>>                                                 
>>    
>> assert_array_almost_equal(w,exact_w[rslice]) 
>>                              
>>  File 
>> "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>> line 
>> 310, in 
>> assert_array_almost_equal 
>>                                                       
>>
>>    header='Arrays are not almost 
>> equal')                                      
>>  File 
>> "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>> line 
>> 295, in 
>> assert_array_compare 
>>                                                           
>>
>>
>>    raise 
>> AssertionError(msg) 
>>                                                 
>> AssertionError: 
>>                                                           
>>     
>>
>> Arrays are not almost 
>> equal 
>>                                                   
>>
>> (mismatch 33.3333333333%)
>> x: array([-0.66992444,  0.48769462,  9.18222713], 
>> dtype=float32)
>> y: array([-0.66992434,  0.48769389,  9.18223045])
>>
>>
>>     
>
> This is a known failure
>
> http://projects.scipy.org/scipy/scipy/ticket/375
>
> However I cannot reproduce the next failure (test_pbdv)
>
> Nils
>
> ======================================================================
>   
>> FAIL: test_pbdv (test_basic.TestCephes)
>> ----------------------------------------------------------------------
>> Traceback (most recent call last):
>>  File 
>> "/usr/lib/python2.5/site-packages/scipy/special/tests/test_basic.py", 
>> line 368, in test_pbdv
>>    assert_equal(cephes.pbdv(1,0),(0.0,0.0))
>>  File 
>> "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>> line 
>> 176, in assert_equal
>>    assert_equal(actual[k], desired[k], 'item=%r\n%s' % 
>> (k,err_msg), 
>> verbose)
>>  File 
>> "/usr/lib/python2.5/site-packages/numpy/testing/utils.py", 
>> line 
>> 183, in assert_equal
>>    raise AssertionError(msg)
>> AssertionError:
>> Items are not equal:
>> item=1
>>
>> ACTUAL: 1.0
>> DESIRED: 0.0
>>
>> I haven't started to have a deep look into this test 
>> code so far.
>> Any ideas?
>>
>> Xavier
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>     
>   
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From dineshbvadhia at hotmail.com  Wed Oct 29 13:42:18 2008
From: dineshbvadhia at hotmail.com (Dinesh B Vadhia)
Date: Wed, 29 Oct 2008 10:42:18 -0700
Subject: [SciPy-user] Sparse matrices and memory usage
Message-ID: <COL103-DS14C369D24594FE158D9D50A3260@phx.gbl>

This question is primarily for Nathan:

We want size the amount of memory required for our application.  Assume a sparse matrix A that consumes N mb of memory.  During a matrix-vector multiplication ie. Ax, how much additional memory is used (for temporary arrays and vectors)?

Dinesh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081029/85ebfd82/attachment.html>

From wnbell at gmail.com  Wed Oct 29 13:53:40 2008
From: wnbell at gmail.com (Nathan Bell)
Date: Wed, 29 Oct 2008 13:53:40 -0400
Subject: [SciPy-user] Sparse matrices and memory usage
In-Reply-To: <COL103-DS14C369D24594FE158D9D50A3260@phx.gbl>
References: <COL103-DS14C369D24594FE158D9D50A3260@phx.gbl>
Message-ID: <d05265cb0810291053p3bfa531ele9a780ef6ecd60dc@mail.gmail.com>

On Wed, Oct 29, 2008 at 1:42 PM, Dinesh B Vadhia
<dineshbvadhia at hotmail.com> wrote:
> This question is primarily for Nathan:
>
> We want size the amount of memory required for our application.  Assume a
> sparse matrix A that consumes N mb of memory.  During a matrix-vector
> multiplication ie. Ax, how much additional memory is used (for temporary
> arrays and vectors)?
>

For CSR and CSC matrices (and a few others), usually only the output
vector (y in y=A*x) needs to be allocated.  However, other formats
that do not provide a matrix-vector method get converted to one that
does.  Also, if the data types of the matrix and vector are not the
same, then one or both are upcast.  For instance, multiplying a
csr_matrix with dtype=int8 by a float64 vector will cause the data
array of the csr_matrix to be upcast to float64 first.

In the future (i.e. SciPy 0.8) we might support mixed types, which
would avoid the upcast.  For now, you should ensure that types match
if you're worried about memory consumption.

-- 
Nathan Bell wnbell at gmail.com
http://graphics.cs.uiuc.edu/~wnbell/


From rowen at u.washington.edu  Wed Oct 29 17:09:32 2008
From: rowen at u.washington.edu (Russell E. Owen)
Date: Wed, 29 Oct 2008 14:09:32 -0700
Subject: [SciPy-user] Is the Mac binary of numpy 1.2.1 usable under MacOS X
	10.4 or 10.3.9?
Message-ID: <rowen-56ECD8.14093229102008@news.gmane.org>

The Mac binary of numpy 1.2.1 claims to be for MacOS X 10.5. Can it be 
used with older versions of the operating system? I'm at 10.4 and have a 
few users still stuck back at 10.3.9.

I can't seem to find this info on the numpy web site, nor in the ReadMe 
of the installer, nor on google.

Note: I use the python.org Mac Python 2.5, not the built-in python.

-- Russell



From wesmckinn at gmail.com  Wed Oct 29 18:23:29 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Wed, 29 Oct 2008 18:23:29 -0400
Subject: [SciPy-user] Enabling NaN-usage in F77 code on Windows
Message-ID: <6c476c8a0810291523n2845000bxdc5486656e91967e@mail.gmail.com>

I'm having some trouble getting NaN's to return from f77 code running under
latest f2py in both g77 and gfortran. I would prefer to use gfortran but
whenever I set a result value = NAN, it comes back to Python as 0. Has
anyone tackled this issue? I am new to using f2py, have been moving along
fine with everything else but ran into this.

Here is a sample function for the rolling mean of a series having this
behavior:

      SUBROUTINE ROLLMEAN(DATA,WINDOW,N,AVE,T)
      INTEGER*8 N, WINDOW
      REAL*8 AVE,DATA(N),T(N)
      INTEGER*8 J
      REAL*8 P,S,S1
C
CF2PY REAL*8 INTENT(IN, COPY) DATA
CF2PY INTEGER*8 INTENT(IN) WINDOW
CF2PY INTEGER*8 INTENT(HIDE), DEPEND(DATA), CHECK(N>=2) :: N = SHAPE(DATA,
0)
CF2PY REAL*8 INTENT(OUT, COPY), DEPEND(N), DIMENSION(N) :: T
CF2PY REAL*8 INTENT(HIDE) :: AVE
C
      S=0.
      S1=0.
      DO J=1,WINDOW
          P=DATA(J)
          S1=S1+P
          T(J)=NAN
      ENDDO
      AVE=S1/WINDOW
      T(WINDOW)=AVE
      DO J=WINDOW+1,N
          P=DATA(J)
          S=DATA(J-WINDOW)
          S1=S1+P-S
          AVE=S1/WINDOW
          T(J)=AVE
      ENDDO
      RETURN
      END


Thanks,
Wes
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081029/2b46d462/attachment.html>

From robert.kern at gmail.com  Wed Oct 29 18:37:12 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 29 Oct 2008 17:37:12 -0500
Subject: [SciPy-user] Enabling NaN-usage in F77 code on Windows
In-Reply-To: <6c476c8a0810291523n2845000bxdc5486656e91967e@mail.gmail.com>
References: <6c476c8a0810291523n2845000bxdc5486656e91967e@mail.gmail.com>
Message-ID: <3d375d730810291537w6ce0a64byd58a46ec4b96bfae@mail.gmail.com>

On Wed, Oct 29, 2008 at 17:23, Wes McKinney <wesmckinn at gmail.com> wrote:
> I'm having some trouble getting NaN's to return from f77 code running under
> latest f2py in both g77 and gfortran. I would prefer to use gfortran but
> whenever I set a result value = NAN, it comes back to Python as 0. Has
> anyone tackled this issue? I am new to using f2py, have been moving along
> fine with everything else but ran into this.

Is NAN a builtin symbol in FORTRAN-77? I don't think it is. I think
what's happening is that the compiler sees you using the name and
implicitly creates a variable for it and initializes it to 0. You will
have to make your own NAN.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From wesmckinn at gmail.com  Wed Oct 29 18:56:07 2008
From: wesmckinn at gmail.com (Wes McKinney)
Date: Wed, 29 Oct 2008 18:56:07 -0400
Subject: [SciPy-user] Enabling NaN-usage in F77 code on Windows
In-Reply-To: <3d375d730810291537w6ce0a64byd58a46ec4b96bfae@mail.gmail.com>
References: <6c476c8a0810291523n2845000bxdc5486656e91967e@mail.gmail.com>
	<3d375d730810291537w6ce0a64byd58a46ec4b96bfae@mail.gmail.com>
Message-ID: <6c476c8a0810291556q61ff147bqa2a76db13b5de3c0@mail.gmail.com>

You would be right about the NAN not being a built-in symbol...any
suggestions for making a NAN? I don't know if there's a a generic way that
will make it back to Python correctly.

On Wed, Oct 29, 2008 at 6:37 PM, Robert Kern <robert.kern at gmail.com> wrote:

> On Wed, Oct 29, 2008 at 17:23, Wes McKinney <wesmckinn at gmail.com> wrote:
> > I'm having some trouble getting NaN's to return from f77 code running
> under
> > latest f2py in both g77 and gfortran. I would prefer to use gfortran but
> > whenever I set a result value = NAN, it comes back to Python as 0. Has
> > anyone tackled this issue? I am new to using f2py, have been moving along
> > fine with everything else but ran into this.
>
> Is NAN a builtin symbol in FORTRAN-77? I don't think it is. I think
> what's happening is that the compiler sees you using the name and
> implicitly creates a variable for it and initializes it to 0. You will
> have to make your own NAN.
>
> --
> Robert Kern
>
> "I have come to believe that the whole world is an enigma, a harmless
> enigma that is made terrible by our own mad attempt to interpret it as
> though it had an underlying truth."
>  -- Umberto Eco
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081029/0dc91654/attachment.html>

From robert.kern at gmail.com  Wed Oct 29 18:57:38 2008
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 29 Oct 2008 17:57:38 -0500
Subject: [SciPy-user] Enabling NaN-usage in F77 code on Windows
In-Reply-To: <6c476c8a0810291556q61ff147bqa2a76db13b5de3c0@mail.gmail.com>
References: <6c476c8a0810291523n2845000bxdc5486656e91967e@mail.gmail.com>
	<3d375d730810291537w6ce0a64byd58a46ec4b96bfae@mail.gmail.com>
	<6c476c8a0810291556q61ff147bqa2a76db13b5de3c0@mail.gmail.com>
Message-ID: <3d375d730810291557j23491aech644e07496a71b73@mail.gmail.com>

On Wed, Oct 29, 2008 at 17:56, Wes McKinney <wesmckinn at gmail.com> wrote:
> You would be right about the NAN not being a built-in symbol...any
> suggestions for making a NAN? I don't know if there's a a generic way that
> will make it back to Python correctly.

inf = 1e200 * 1e200
nan = inf * 0

Most likely.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless
enigma that is made terrible by our own mad attempt to interpret it as
though it had an underlying truth."
  -- Umberto Eco


From millman at berkeley.edu  Thu Oct 30 00:20:46 2008
From: millman at berkeley.edu (Jarrod Millman)
Date: Wed, 29 Oct 2008 21:20:46 -0700
Subject: [SciPy-user] Is the Mac binary of numpy 1.2.1 usable under
	MacOS X 10.4 or 10.3.9?
In-Reply-To: <rowen-56ECD8.14093229102008@news.gmane.org>
References: <rowen-56ECD8.14093229102008@news.gmane.org>
Message-ID: <c7009a550810292120s6cfb8aa5l1affbbe20f57f08a@mail.gmail.com>

On Wed, Oct 29, 2008 at 2:09 PM, Russell E. Owen <rowen at u.washington.edu> wrote:
> The Mac binary of numpy 1.2.1 claims to be for MacOS X 10.5. Can it be
> used with older versions of the operating system? I'm at 10.4 and have a
> few users still stuck back at 10.3.9.

I built the dmg and believe that it should work with 10.3.9 or
greater, but I haven't tested it.  If you have a test system, try it
and run the tests.  Please report back with the results.  If it works,
I will update the release notes.  If not, we can try and track down
the problem and build new binaries.

> Note: I use the python.org Mac Python 2.5, not the built-in python.

Good, the NumPy binary requires that you use the python.org binary.

-- 
Jarrod Millman
Computational Infrastructure for Research Labs
10 Giannini Hall, UC Berkeley
phone: 510.643.4014
http://cirl.berkeley.edu/


From tgrav at mac.com  Thu Oct 30 00:36:37 2008
From: tgrav at mac.com (Tommy Grav)
Date: Thu, 30 Oct 2008 00:36:37 -0400
Subject: [SciPy-user] Is the Mac binary of numpy 1.2.1 usable
 under	MacOS X 10.4 or 10.3.9?
In-Reply-To: <c7009a550810292120s6cfb8aa5l1affbbe20f57f08a@mail.gmail.com>
References: <rowen-56ECD8.14093229102008@news.gmane.org>
	<c7009a550810292120s6cfb8aa5l1affbbe20f57f08a@mail.gmail.com>
Message-ID: <7445DA88-4EB3-4934-A41B-82E312B28812@mac.com>

On Oct 30, 2008, at 12:20 AM, Jarrod Millman wrote:
>
> Good, the NumPy binary requires that you use the python.org binary.

It also works fine with

ActivePython 2.5.2.2 (ActiveState Software Inc.) based on
Python 2.5.2 (r252:60911, Mar 27 2008, 17:40:23)
[GCC 4.0.1 (Apple Computer, Inc. build 5250)] on darwin
Type "help", "copyright", "credits" or "license" for more information.

on a Intel Macbook running 10.5.5. All test pass except one knownfail.

Cheers
   Tommy


From david at ar.media.kyoto-u.ac.jp  Thu Oct 30 00:35:56 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Thu, 30 Oct 2008 13:35:56 +0900
Subject: [SciPy-user] run=2339 errors=0 failures=3 on ibex
In-Reply-To: <4908941F.6000901@gmail.com>
References: <490851E7.1020706@gmail.com> <web-113448791@uni-stuttgart.de>
	<4908941F.6000901@gmail.com>
Message-ID: <490939AC.9030808@ar.media.kyoto-u.ac.jp>

Xavier Gnata wrote:
> Ok so there is a simple fix which is fully correct:
>
> -assert_array_almost_equal(x, y, decimal=6, err_msg='', verbose=True)
>
> +assert_array_almost_equal(x, y, decimal=5, err_msg='', verbose=True)
>   

The obvious question is does that fix the issue or does it only hide the
problem ? 5 decimals is relatively poor, but float32 can only be
expected to have 7 decimals anyway. And assuming the failure is caused
by different BLAS/LAPACK implementations, it is not rare to see
relatively significant differences between two given BLAS/LAPACK (even
same version but different compilers; I have never seen a BLAS/LAPACK
passing all the netlib LAPACK tests, for example: g77 and gfortran break
them at different places, for example).

Did you use g77 before intrepid (intrepid finally use gfortran as the
default ABI for fortran) ?

cheers,

David


From fredmfp at gmail.com  Thu Oct 30 04:47:32 2008
From: fredmfp at gmail.com (fred)
Date: Thu, 30 Oct 2008 09:47:32 +0100
Subject: [SciPy-user] Enabling NaN-usage in F77 code on Windows
In-Reply-To: <3d375d730810291557j23491aech644e07496a71b73@mail.gmail.com>
References: <6c476c8a0810291523n2845000bxdc5486656e91967e@mail.gmail.com>	<3d375d730810291537w6ce0a64byd58a46ec4b96bfae@mail.gmail.com>	<6c476c8a0810291556q61ff147bqa2a76db13b5de3c0@mail.gmail.com>
	<3d375d730810291557j23491aech644e07496a71b73@mail.gmail.com>
Message-ID: <490974A4.3040902@gmail.com>

Robert Kern a ?crit :
> On Wed, Oct 29, 2008 at 17:56, Wes McKinney <wesmckinn at gmail.com> wrote:
>> You would be right about the NAN not being a built-in symbol...any
>> suggestions for making a NAN? I don't know if there's a a generic way that
>> will make it back to Python correctly.
> 
> inf = 1e200 * 1e200
> nan = inf * 0
> 
> Most likely.

I use:
foo = -1
nan = sqrt(foo)


Cheers,

-- 
Fred


From rowen at u.washington.edu  Thu Oct 30 12:51:36 2008
From: rowen at u.washington.edu (Russell E. Owen)
Date: Thu, 30 Oct 2008 09:51:36 -0700
Subject: [SciPy-user] Is the Mac binary of numpy 1.2.1 usable under
	MacOS X 10.4 or 10.3.9?
References: <rowen-56ECD8.14093229102008@news.gmane.org>
	<c7009a550810292120s6cfb8aa5l1affbbe20f57f08a@mail.gmail.com>
Message-ID: <rowen-B31C4C.09513630102008@news.gmane.org>

In article 
<c7009a550810292120s6cfb8aa5l1affbbe20f57f08a at mail.gmail.com>,
 "Jarrod Millman" <millman at berkeley.edu> wrote:

> On Wed, Oct 29, 2008 at 2:09 PM, Russell E. Owen <rowen at u.washington.edu> 
> wrote:
> > The Mac binary of numpy 1.2.1 claims to be for MacOS X 10.5. Can it be
> > used with older versions of the operating system? I'm at 10.4 and have a
> > few users still stuck back at 10.3.9.
> 
> I built the dmg and believe that it should work with 10.3.9 or
> greater, but I haven't tested it.  If you have a test system, try it
> and run the tests.  Please report back with the results.  If it works,
> I will update the release notes.  If not, we can try and track down
> the problem and build new binaries.
> 
> > Note: I use the python.org Mac Python 2.5, not the built-in python.
> 
> Good, the NumPy binary requires that you use the python.org binary.

On MacOS X 10.4.11 (I don't have access to 10.3.9) and after installing 
"nose", numpy.test() runs fine for awhile, then prints the appended text 
and exits python. Two puzzles:
- What do the initial set of messages mean?
- Why does it crash out of Python if it cannot find a fortran compiler? 
I can install one if necessary, but I'm not very keen to have one around 
just to run tests, especially since the fortran world is so terribly 
fragmented right now.
This is with Python 2.5.2 from python.org.

-- Russell

Not implemented: Defined_Binary_Op
Not implemented: Defined_Binary_Op
Defined_Operator not defined used by Generic_Spec
Needs match implementation: Allocate_Stmt
Needs match implementation: Associate_Construct
Needs match implementation: Backspace_Stmt
Needs match implementation: Block_Data
Needs match implementation: Case_Construct
Needs match implementation: Case_Selector
Needs match implementation: Case_Stmt
Needs match implementation: Control_Edit_Desc
Needs match implementation: Data_Component_Def_Stmt
Needs match implementation: Data_Implied_Do
Needs match implementation: Data_Stmt
Needs match implementation: Data_Stmt_Set
Needs match implementation: Deallocate_Stmt
Needs match implementation: Derived_Type_Def
Needs match implementation: Endfile_Stmt
Needs match implementation: Entry_Stmt
Needs match implementation: Enum_Def
Needs match implementation: Flush_Stmt
Needs match implementation: Forall_Construct
Needs match implementation: Forall_Header
Needs match implementation: Forall_Triplet_Spec
Needs match implementation: Format_Item
Needs match implementation: Function_Stmt
Needs match implementation: Generic_Binding
Needs match implementation: Generic_Spec
Needs match implementation: Implicit_Part
Needs match implementation: Inquire_Stmt
Needs match implementation: Interface_Block
Needs match implementation: Interface_Body
Needs match implementation: Interface_Stmt
Needs match implementation: Internal_Subprogram_Part
Needs match implementation: Io_Implied_Do
Needs match implementation: Io_Implied_Do_Control
Needs match implementation: Main_Program
Needs match implementation: Module
Needs match implementation: Module_Subprogram_Part
Needs match implementation: Namelist_Stmt
Needs match implementation: Pointer_Assignment_Stmt
Needs match implementation: Position_Edit_Desc
Needs match implementation: Proc_Attr_Spec
Needs match implementation: Proc_Component_Def_Stmt
Needs match implementation: Procedure_Declaration_Stmt
Needs match implementation: Procedure_Stmt
Needs match implementation: Read_Stmt
Needs match implementation: Rewind_Stmt
Needs match implementation: Select_Type_Construct
Needs match implementation: Select_Type_Stmt
Needs match implementation: Specific_Binding
Needs match implementation: Target_Stmt
Needs match implementation: Type_Bound_Procedure_Part
Needs match implementation: Where_Construct
-----
Nof match implementation needs: 51 out of 224
Nof tests needs: 224 out of 224
Total number of classes: 529
-----
No module named test_derived_scalar_ext , recompiling 
test_derived_scalar_ext.
Parsing '/tmp/tmpYXkcLK.f90'..
Generating interface for test_derived_scalar_ext Subroutine: 
f2py_test_derived_scalar_ext_foo
Generating interface for test_derived_scalar_ext.myt: f2py_type_myt_32
Generating interface for Integer: npy_int32
Generating interface for test_derived_scalar_ext Subroutine: 
f2py_test_derived_scalar_ext_f2pywrap_foo2
setup arguments: ' build_ext --build-temp tmp/ext_temp --build-lib tmp 
build_clib --build-temp tmp/clib_temp --build-clib tmp/clib_clib'
running build_ext
running build_src
building library "test_derived_scalar_ext_fortran_f2py" sources
building library "test_derived_scalar_ext_f_wrappers_f2py" sources
building extension "test_derived_scalar_ext" sources
running build_clib
customize UnixCCompiler
customize UnixCCompiler using build_clib
customize NAGFCompiler
Could not locate executable f95
customize AbsoftFCompiler
Could not locate executable f90
Could not locate executable f77
customize IBMFCompiler
Could not locate executable xlf90
Could not locate executable xlf
customize IntelFCompiler
Could not locate executable ifort
Could not locate executable ifc
customize GnuFCompiler
Could not locate executable g77
customize Gnu95FCompiler
Could not locate executable gfortran
customize G95FCompiler
Could not locate executable g95
don't know how to compile Fortran code on platform 'posix'
building 'test_derived_scalar_ext_fortran_f2py' library
error: library test_derived_scalar_ext_fortran_f2py has Fortran sources 
but no Fortran compiler found



From soren.skou.nielsen at gmail.com  Thu Oct 30 14:24:02 2008
From: soren.skou.nielsen at gmail.com (=?ISO-8859-1?Q?S=F8ren_Nielsen?=)
Date: Thu, 30 Oct 2008 19:24:02 +0100
Subject: [SciPy-user] Weave extension with lots of input and returns
Message-ID: <de1aa14a0810301124m4cf300cdn1cd7f1864297883a@mail.gmail.com>

Hi,

I'm trying to make a weave python extension to use in my program. I already
did it in inline, but that doesn't work with py2exe (needs compiler), so I'm
creating extensions instead.

Heres the problem. Inline took care of everything, ext_tools obviously
doesn't. how do I make a function that accepts like 10 inputs of numpy
arrays and different constants, and returns 3 arrays after processing?

All examples in the package are very simple and mostly deal with a single
Int or a single PyObject.. Should I declare them all as PyObjects?

So that:

PyObject MyFunction(PyObject* a, PyObject* b, etc etc.., PyObject* k) {

   C processing code
}

and how do I then return three arrays?

Any help is appreciated!

Thanks,
Soren
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081030/ed6ffb18/attachment.html>

From xavier.gnata at gmail.com  Thu Oct 30 16:12:57 2008
From: xavier.gnata at gmail.com (Xavier Gnata)
Date: Thu, 30 Oct 2008 21:12:57 +0100
Subject: [SciPy-user] run=2339 errors=0 failures=3 on ibex
In-Reply-To: <490939AC.9030808@ar.media.kyoto-u.ac.jp>
References: <490851E7.1020706@gmail.com>
	<web-113448791@uni-stuttgart.de>	<4908941F.6000901@gmail.com>
	<490939AC.9030808@ar.media.kyoto-u.ac.jp>
Message-ID: <490A1549.2020109@gmail.com>

David Cournapeau wrote:
> Xavier Gnata wrote:
>   
>> Ok so there is a simple fix which is fully correct:
>>
>> -assert_array_almost_equal(x, y, decimal=6, err_msg='', verbose=True)
>>
>> +assert_array_almost_equal(x, y, decimal=5, err_msg='', verbose=True)
>>   
>>     
>
> The obvious question is does that fix the issue or does it only hide the
> problem ? 5 decimals is relatively poor, but float32 can only be
> expected to have 7 decimals anyway. And assuming the failure is caused
> by different BLAS/LAPACK implementations, it is not rare to see
> relatively significant differences between two given BLAS/LAPACK (even
> same version but different compilers; I have never seen a BLAS/LAPACK
> passing all the netlib LAPACK tests, for example: g77 and gfortran break
> them at different places, for example).
>
>   
Ok my fault. It was not a good idea to "fix" it that way.
So first I can try to implement the same computation in C/whatever using 
float32 and see how many correct digits we get.
If it "breaks" on some systems, IMHO it means that the check is too 
stringent.

> Did you use g77 before intrepid (intrepid finally use gfortran as the
> default ABI for fortran) ?
>
>   
Full gfortran. No g77.

So what should be the way forward?  Should we involve gfortran/lapack 
guys asking for how many digits we should get?
Should we move from an error to a warning (because the result may only 
be not as accurate as expected).

As you know, I really would like to see scipy compiling and running the 
test suite smoothly :)
It is very important to "sell" scipy to people (people I know at least ;)).

Cheers,
Xavier


> cheers,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>   



From cdcasey at gmail.com  Thu Oct 30 19:48:37 2008
From: cdcasey at gmail.com (chris)
Date: Thu, 30 Oct 2008 18:48:37 -0500
Subject: [SciPy-user] scipy 0.6.0 build failing
In-Reply-To: <5b8d13220810162216x96b9c8ek3648fed2724fe999@mail.gmail.com>
References: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>
	<5b8d13220810162216x96b9c8ek3648fed2724fe999@mail.gmail.com>
Message-ID: <ead057790810301648y62918d4at8e9be93f7867f7bf@mail.gmail.com>

I'm using g77 3.2.3 on RedHat 3. Is there a particular version of
g77/gcc I should updrage to? A fresh compile of 3.2.3 gave the same
error.

-Chris

On Fri, Oct 17, 2008 at 12:16 AM, David Cournapeau <cournape at gmail.com> wrote:
> On Tue, Oct 14, 2008 at 3:09 AM, chris <cdcasey at gmail.com> wrote:
>> I'm trying to build scipy 0.6.0 on RHEL 3, and am getting the following failure:
>>
>>
>> g77:f77: scipy/fftpack/dfftpack/zfftf1.f
>> /tmp/cceGs6VT.s: Assembler messages:
>> /tmp/cceGs6VT.s:598: Error: suffix or operands invalid for `movd'
>
> Your version of g77 has a bug, and generate invalid machine
> instructions when the -msse option is used. You should update your g77
> if possible,
>
> David
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>


From cdcasey at gmail.com  Thu Oct 30 20:33:54 2008
From: cdcasey at gmail.com (chris)
Date: Thu, 30 Oct 2008 19:33:54 -0500
Subject: [SciPy-user] scipy 0.6.0 build failing
In-Reply-To: <ead057790810301648y62918d4at8e9be93f7867f7bf@mail.gmail.com>
References: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>
	<5b8d13220810162216x96b9c8ek3648fed2724fe999@mail.gmail.com>
	<ead057790810301648y62918d4at8e9be93f7867f7bf@mail.gmail.com>
Message-ID: <ead057790810301733w36be4c2k1768c6864888718@mail.gmail.com>

And maybe more importantly, will I have problems if I use g77 3.4.x
and gcc 3.2.x?

On Thu, Oct 30, 2008 at 6:48 PM, chris <cdcasey at gmail.com> wrote:
> I'm using g77 3.2.3 on RedHat 3. Is there a particular version of
> g77/gcc I should updrage to? A fresh compile of 3.2.3 gave the same
> error.
>
> -Chris
>
> On Fri, Oct 17, 2008 at 12:16 AM, David Cournapeau <cournape at gmail.com> wrote:
>> On Tue, Oct 14, 2008 at 3:09 AM, chris <cdcasey at gmail.com> wrote:
>>> I'm trying to build scipy 0.6.0 on RHEL 3, and am getting the following failure:
>>>
>>>
>>> g77:f77: scipy/fftpack/dfftpack/zfftf1.f
>>> /tmp/cceGs6VT.s: Assembler messages:
>>> /tmp/cceGs6VT.s:598: Error: suffix or operands invalid for `movd'
>>
>> Your version of g77 has a bug, and generate invalid machine
>> instructions when the -msse option is used. You should update your g77
>> if possible,
>>
>> David
>> _______________________________________________
>> SciPy-user mailing list
>> SciPy-user at scipy.org
>> http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>


From tjhnson at gmail.com  Thu Oct 30 20:43:45 2008
From: tjhnson at gmail.com (T J)
Date: Thu, 30 Oct 2008 17:43:45 -0700
Subject: [SciPy-user] swig c++ class storing pointer to array
Message-ID: <b02c60870810301743x7a884901ufd34e7992fd4a334@mail.gmail.com>

Hi,

I'm new to using swig and am struggling with arrays and python.  I
have created a C++ class which stores a pointer to an array as a
member.  Using the numpy.i interface file, I would like to be able to
pass in arrays from Python (without copying) to the constructor.  The
end goal is to be able to pass in an numpy array, use the C++ class to
do some operations on the array, create some new arrays, and then be
able to access those arrays (as they are updated) from within Python.

To get started, I've created a very basic file with implements the
storing of the arrays, and here is how I used it:

>>> import testme
>>> x = testme.MyClass([1,2,3])
>>> x.printme()
1.5865e-268
1.08557e-269
1.08557e-269
>>> a = numpy.array([1,2,3.])
>>> x = testme.MyClass(a)
>>> x.printme()
4.18205e-62
3.60756e-313
1.60847e-268

So I am getting uninitialized values.  When I look at
_wrap_new_MyClass in the wrapper file created by swig, I can insert:

  result->printme()

before "return resultobj", and the proper values are printed.  So I'm
not understanding why this isn't working...or what pointer is being
stored in the class.  I've attached all the necessary files.

Curiously, the following does seem to work:

>>> x = numpy.array([.5,.3,2.])
>>> y = numpy.array(x)
>>> z = testme.MyClass(y)
>>> z.printme()
0.5
0.3
2

Can someone explain this?
-------------- next part --------------
A non-text attachment was scrubbed...
Name: testme.i
Type: application/octet-stream
Size: 224 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081030/74faabb7/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: numpy.i
Type: application/octet-stream
Size: 56154 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081030/74faabb7/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: testme.h
Type: text/x-chdr
Size: 163 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081030/74faabb7/attachment.h>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: testme.cpp
Type: text/x-c++src
Size: 249 bytes
Desc: not available
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081030/74faabb7/attachment.cpp>

From david at ar.media.kyoto-u.ac.jp  Fri Oct 31 00:28:43 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 31 Oct 2008 13:28:43 +0900
Subject: [SciPy-user] scipy 0.6.0 build failing
In-Reply-To: <ead057790810301648y62918d4at8e9be93f7867f7bf@mail.gmail.com>
References: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>	<5b8d13220810162216x96b9c8ek3648fed2724fe999@mail.gmail.com>
	<ead057790810301648y62918d4at8e9be93f7867f7bf@mail.gmail.com>
Message-ID: <490A897B.9050001@ar.media.kyoto-u.ac.jp>

chris wrote:
> I'm using g77 3.2.3 on RedHat 3. Is there a particular version of
> g77/gcc I should updrage to? A fresh compile of 3.2.3 gave the same
> error.
>   

Yep, that's expected: the upstream version of 3.2.3 is the culprit if
that was not clear. The exact problem is the use of the -msse2 flag; I
believe that if you remove this flag from the build, it should build
correctly. That would certainly be easier to try than compiling your own
compiler :)

Now, unfortunately, there is no easy way to modify those compiler flags
without modifying the source code. You could remove lines 268-270 (the
exact numbers may be difference depending on the version of numpy you
are using), in the file numpy/distutils/fcompiler/gnu.py:

 if gnu_ver > '3.2.2':
            if cpu.has_sse2(): opt.append('-msse2')
            if cpu.has_sse(): opt.append('-msse')

After the modification, remove the build directory of numpy entirely,
and start the build from scratch.

cheers,

David


From kartita at gmail.com  Fri Oct 31 00:52:20 2008
From: kartita at gmail.com (Kimberly Artita)
Date: Thu, 30 Oct 2008 23:52:20 -0500
Subject: [SciPy-user] f2py "Segmentation fault"-revisited, please help
Message-ID: <ff948bef0810302152q146e8fbcp7d8ce045c96eea14@mail.gmail.com>

Hi,

Can someone please tell me why I keep getting a segmentation fault?

my fortran script (gfortran_test.f90):
subroutine readin_test

    implicit none

    character(len=4) :: title (60)
    character (len=13) :: bigsub, sbsout, rchout, rsvout, lwqout, wtrout
    open (2,file="gfortran.txt", delim='none')
    print *, "title"
    read (2,5100) title
    print *, title
    read (2,5000) bigsub, sbsout, rchout, rsvout, lwqout, wtrout

    print *, "bigsub, sbsout, rchout, rsvout, lwqout, wtrout"
    print *, bigsub, sbsout, rchout, rsvout, lwqout, wtrout
    close(2)

    5100 format (20a4)
    5000 format (6a)

end subroutine readin_test

my python script (gfortran_test.py):
import gfortran_test

gfortran_test.readin_test()

my text file (gfortran.txt):
General Input/Output section (file.cio):           Thu Mar 13 17:32:19 2008
AVSWAT2000 - SWAT interface MDL


   basins.bsb   basins.sbs   basins.rch   basins.rsv   basins.lqo
basins.wtr


using this version of gfortran: i686-pc-linux-gnu-4.1.2 with either
numpy-1.0.4-r2 or numpy-1.2.0

I can compile gfortran_test.f90 as a standalone program and it works!

BUT, when I call it as a subroutine from python using f2py, it fails!
I type: f2py --fcompiler=gfortran -c -m gfortran_test gfortran_test.f90


Why??????

-- 
Kimberly S. Artita
PhD Intern, CDM
Graduate Student, Engineering Science
Southern Illinois University Carbondale
Carbondale, Illinois 62901-6603
(618)-528-0349
e-mail: kartita at gmail.com, kartita at siu.edu
web: http://civil.engr.siu.edu/GraduateStudents/artita/index.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081030/896c115a/attachment.html>

From david at ar.media.kyoto-u.ac.jp  Fri Oct 31 00:52:59 2008
From: david at ar.media.kyoto-u.ac.jp (David Cournapeau)
Date: Fri, 31 Oct 2008 13:52:59 +0900
Subject: [SciPy-user] f2py "Segmentation fault"-revisited, please help
In-Reply-To: <ff948bef0810302152q146e8fbcp7d8ce045c96eea14@mail.gmail.com>
References: <ff948bef0810302152q146e8fbcp7d8ce045c96eea14@mail.gmail.com>
Message-ID: <490A8F2B.9070403@ar.media.kyoto-u.ac.jp>

Kimberly Artita wrote:
> Hi,
>
> Can someone please tell me why I keep getting a segmentation fault?
>
> my fortran script (gfortran_test.f90):
> subroutine readin_test
>    
>     implicit none
>    
>     character(len=4) :: title (60)
>     character (len=13) :: bigsub, sbsout, rchout, rsvout, lwqout, wtrout
>     open (2,file="gfortran.txt", delim='none')
>     print *, "title"
>     read (2,5100) title
>     print *, title   
>     read (2,5000) bigsub, sbsout, rchout, rsvout, lwqout, wtrout
>    
>     print *, "bigsub, sbsout, rchout, rsvout, lwqout, wtrout"
>     print *, bigsub, sbsout, rchout, rsvout, lwqout, wtrout
>     close(2)
>    
>     5100 format (20a4)
>     5000 format (6a)
>    

I don't know about the exact problem, but C/Fortran mixing is already
quite error prone and has many warts, and file IO is even worse (because
the C runtime and the fortran runtimes must cooperate in the same
process, and they generally don't cooperate well). You should avoid it
if you can.

> I can compile gfortran_test.f90 as a standalone program and it works!

Can you try with a main written in C (numpy is in C) ? I would not be
surprised if that's the issue,

cheers,

David


From markbak at gmail.com  Fri Oct 31 05:49:08 2008
From: markbak at gmail.com (Mark Bakker)
Date: Fri, 31 Oct 2008 10:49:08 +0100
Subject: [SciPy-user] overlapping polygons
Message-ID: <6946b9500810310249i10b8d62fy3ace317f2921391b@mail.gmail.com>

Hello list -

I am now using the shapely Python package to compute the intersection or
difference of two closed polygons.

Does scipy (or numpy) have similar abilities?

Any other suggestions? Not that I don't like shapely, it just means one
additional dependency for my code.

Thanks, Mark
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081031/7af901ff/attachment.html>

From gnurser at googlemail.com  Fri Oct 31 06:20:07 2008
From: gnurser at googlemail.com (George Nurser)
Date: Fri, 31 Oct 2008 10:20:07 +0000
Subject: [SciPy-user] f2py "Segmentation fault"-revisited, please help
In-Reply-To: <ff948bef0810302152q146e8fbcp7d8ce045c96eea14@mail.gmail.com>
References: <ff948bef0810302152q146e8fbcp7d8ce045c96eea14@mail.gmail.com>
Message-ID: <1d1e6ea70810310320y108d0edr1af274637f7ae774@mail.gmail.com>

Hi,

2008/10/31 Kimberly Artita <kartita at gmail.com>:
> Hi,
>
> Can someone please tell me why I keep getting a segmentation fault?
[cut]

You need to compile with fcompiler=gnu95

>
> I type: f2py --fcompiler=gfortran -c -m gfortran_test gfortran_test.f90

Do
f2py --fcompiler=gnu95 -c -m gfortran_test gfortran_test.f90

It worked fine for me (gfortran 4.3.2, Numpy 1.3.0.dev5867, Mac OS X)

HTH, George Nurser.


From stefan at sun.ac.za  Fri Oct 31 08:18:24 2008
From: stefan at sun.ac.za (=?ISO-8859-1?Q?St=E9fan_van_der_Walt?=)
Date: Fri, 31 Oct 2008 14:18:24 +0200
Subject: [SciPy-user] overlapping polygons
In-Reply-To: <6946b9500810310249i10b8d62fy3ace317f2921391b@mail.gmail.com>
References: <6946b9500810310249i10b8d62fy3ace317f2921391b@mail.gmail.com>
Message-ID: <9457e7c80810310518h7a57e425j9ee9beb80ba64b97@mail.gmail.com>

Hi Mark

These capabilities are not available in SciPy or NumPy.  I have BSD
polygon clipping code for the case where you intersect one rectangle
and another convex polygon.  If you need to calculate the intersection
between two arbitrary polygons, you can also use Joerg Raedler's gpc
wrapper (it's a very light-weight dependency):

http://bazaar.launchpad.net/%7Estefanv/supreme/main/files/180?file_id=Polygon1.16-20060502184912-b1a90b5a4870e7ba

That's one ugly URL.  Sorry.

Cheers
St?fan

2008/10/31 Mark Bakker <markbak at gmail.com>:
> Hello list -
>
> I am now using the shapely Python package to compute the intersection or
> difference of two closed polygons.
>
> Does scipy (or numpy) have similar abilities?
>
> Any other suggestions? Not that I don't like shapely, it just means one
> additional dependency for my code.
>
> Thanks, Mark


From soren.skou.nielsen at gmail.com  Fri Oct 31 10:43:14 2008
From: soren.skou.nielsen at gmail.com (=?ISO-8859-1?Q?S=F8ren_Nielsen?=)
Date: Fri, 31 Oct 2008 15:43:14 +0100
Subject: [SciPy-user] Blitz with ext_tools
Message-ID: <de1aa14a0810310743t1687b482p47e3a1cf3c415d04@mail.gmail.com>

Hi,

Is there a way I can use blitz with ext_tools? so that I can refer to arrays
like a(x,y) in the c code??

Soren
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081031/91d70701/attachment.html>

From cdcasey at gmail.com  Fri Oct 31 12:40:24 2008
From: cdcasey at gmail.com (chris)
Date: Fri, 31 Oct 2008 11:40:24 -0500
Subject: [SciPy-user] scipy 0.6.0 build failing
In-Reply-To: <490A897B.9050001@ar.media.kyoto-u.ac.jp>
References: <ead057790810131109w15ef79cah4757579bb9643960@mail.gmail.com>
	<5b8d13220810162216x96b9c8ek3648fed2724fe999@mail.gmail.com>
	<ead057790810301648y62918d4at8e9be93f7867f7bf@mail.gmail.com>
	<490A897B.9050001@ar.media.kyoto-u.ac.jp>
Message-ID: <ead057790810310940r41c17092r8cdb80f0acb1fe29@mail.gmail.com>

On Thu, Oct 30, 2008 at 11:28 PM, David Cournapeau
<david at ar.media.kyoto-u.ac.jp> wrote:
> chris wrote:
>> I'm using g77 3.2.3 on RedHat 3. Is there a particular version of
>> g77/gcc I should updrage to? A fresh compile of 3.2.3 gave the same
>> error.
>>
>
> Yep, that's expected: the upstream version of 3.2.3 is the culprit if
> that was not clear. The exact problem is the use of the -msse2 flag; I
> believe that if you remove this flag from the build, it should build
> correctly. That would certainly be easier to try than compiling your own
> compiler :)
>
> Now, unfortunately, there is no easy way to modify those compiler flags
> without modifying the source code. You could remove lines 268-270 (the
> exact numbers may be difference depending on the version of numpy you
> are using), in the file numpy/distutils/fcompiler/gnu.py:
>
>  if gnu_ver > '3.2.2':
>            if cpu.has_sse2(): opt.append('-msse2')
>            if cpu.has_sse(): opt.append('-msse')
>
> After the modification, remove the build directory of numpy entirely,
> and start the build from scratch.
>
If only the -msse2 flag is causing the problem, would it be sufficient
to remove line 269? Or is there a reason they both have to be removed?

-Chris


From kwmsmith at gmail.com  Fri Oct 31 13:09:59 2008
From: kwmsmith at gmail.com (Kurt Smith)
Date: Fri, 31 Oct 2008 12:09:59 -0500
Subject: [SciPy-user] Explanation of different edge modes in scipy.ndimage
Message-ID: <c266a7930810311009l631668b4h4f219e8afc58ea0c@mail.gmail.com>

Hello,

I'm doing some gaussian filtering of periodic 2D arrays using
scipy.ndimage.gaussian_filter.  There is a 'mode' argument that is set to
'reflect' by default.  In _ni_support.py:34 there is a conversion function,
'_extend_mode_to_code' that gives the different modes available.  For
periodic data I believe I should use 'wrap', but I'm interested to know what
the other modes mean, esp the difference between 'reflect' and 'mirror'.
For the record, the modes defined are 'nearest', 'wrap', 'reflect',
'mirror', and 'constant'.  For future reference, is there a place where
these arguments are documented?

Thanks,

Kurt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081031/80d706df/attachment.html>

From w.kejia at gmail.com  Fri Oct 31 13:21:38 2008
From: w.kejia at gmail.com (Wu, Kejia)
Date: Fri, 31 Oct 2008 10:21:38 -0700
Subject: [SciPy-user] About Random Number Generation
Message-ID: <1225473698.7737.2.camel@localhost>

Hi all,

I tried the example code here:
http://numpy.scipy.org/numpydoc/numpy-20.html#71863
But failed:
--------------------------------------
rng.py, line 5, in <module>
    import RNG
ImportError: No module named RNG
--------------------------------------

Any suggestion? Thanks at first.

Also, can any body tell me whether the random number algorithm in RNG
package is a pseudorandom one or a real-random one? And is there an
available implementation for Monte Carlo method in NumPy?

Thanks a lot for any reply.

Regards,
Kejia






From kartita at gmail.com  Fri Oct 31 13:34:53 2008
From: kartita at gmail.com (Kimberly Artita)
Date: Fri, 31 Oct 2008 12:34:53 -0500
Subject: [SciPy-user] f2py "Segmentation fault"-revisited, please help
In-Reply-To: <ff948bef0810302152q146e8fbcp7d8ce045c96eea14@mail.gmail.com>
References: <ff948bef0810302152q146e8fbcp7d8ce045c96eea14@mail.gmail.com>
Message-ID: <ff948bef0810311034h3f52f4a1uc7a716bf9249dd5d@mail.gmail.com>

Tried it on a different machine (numpy-1.2.0 and gcc-4.3.2 on linux)

Either way (--fcompiler=gnu95 or gfortran) gives a segfault
The output says "General", then segfaults.
It is reading the space as a delimiter, even though I specify delim='none'

My laptop and the desktop used above run gentoo.  A third desktop using
ubuntu with gcc-4.3.2 and numpy-1.2.0 works fine.  What gives?

Kim

On Fri, Oct 31, 2008 at 5:20 AM, George Nurser <gnurser at googlemail.com>wrote:

> Hi,
>
> 2008/10/31 Kimberly Artita <kartita at gmail.com>:
> > Hi,
> >
> > Can someone please tell me why I keep getting a segmentation fault?
> [cut]
>
> You need to compile with fcompiler=gnu95
>
> >
> > I type: f2py --fcompiler=gfortran -c -m gfortran_test gfortran_test.f90
>
> Do
> f2py --fcompiler=gnu95 -c -m gfortran_test gfortran_test.f90
>
> It worked fine for me (gfortran 4.3.2, Numpy 1.3.0.dev5867, Mac OS X)
>
> HTH, George Nurser.
>  _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.org
> http://projects.scipy.org/mailman/listinfo/scipy-user
>



On Thu, Oct 30, 2008 at 11:52 PM, Kimberly Artita <kartita at gmail.com> wrote:

> Hi,
>
> Can someone please tell me why I keep getting a segmentation fault?
>
> my fortran script (gfortran_test.f90):
> subroutine readin_test
>
>     implicit none
>
>     character(len=4) :: title (60)
>     character (len=13) :: bigsub, sbsout, rchout, rsvout, lwqout, wtrout
>     open (2,file="gfortran.txt", delim='none')
>     print *, "title"
>     read (2,5100) title
>     print *, title
>     read (2,5000) bigsub, sbsout, rchout, rsvout, lwqout, wtrout
>
>     print *, "bigsub, sbsout, rchout, rsvout, lwqout, wtrout"
>     print *, bigsub, sbsout, rchout, rsvout, lwqout, wtrout
>     close(2)
>
>     5100 format (20a4)
>     5000 format (6a)
>
> end subroutine readin_test
>
> my python script (gfortran_test.py):
> import gfortran_test
>
> gfortran_test.readin_test()
>
> my text file (gfortran.txt):
> General Input/Output section (file.cio):           Thu Mar 13 17:32:19 2008
> AVSWAT2000 - SWAT interface MDL
>
>
>    basins.bsb   basins.sbs   basins.rch   basins.rsv   basins.lqo
> basins.wtr
>
>
> using this version of gfortran: i686-pc-linux-gnu-4.1.2 with either
> numpy-1.0.4-r2 or numpy-1.2.0
>
> I can compile gfortran_test.f90 as a standalone program and it works!
>
> BUT, when I call it as a subroutine from python using f2py, it fails!
> I type: f2py --fcompiler=gfortran -c -m gfortran_test gfortran_test.f90
>
>
> Why??????
>
> --
> Kimberly S. Artita
> PhD Intern, CDM
> Graduate Student, Engineering Science
> Southern Illinois University Carbondale
> Carbondale, Illinois 62901-6603
> (618)-528-0349
> e-mail: kartita at gmail.com, kartita at siu.edu
> web: http://civil.engr.siu.edu/GraduateStudents/artita/index.html
>



-- 
Kimberly S. Artita
PhD Intern, CDM
Graduate Student, Engineering Science
Southern Illinois University Carbondale
Carbondale, Illinois 62901-6603
(618)-528-0349
e-mail: kartita at gmail.com, kartita at siu.edu
web: http://civil.engr.siu.edu/GraduateStudents/artita/index.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20081031/2ecb3baf/attachment.html>

From dwf at cs.toronto.edu  Fri Oct 31 14:31:55 2008
From: dwf at cs.toronto.edu (David Warde-Farley)
Date: Fri, 31 Oct 2008 14:31:55 -0400
Subject: [SciPy-user] About Random Number Generation
In-Reply-To: <1225473698.7737.2.camel@localhost>
References: <1225473698.7737.2.camel@localhost>
Message-ID: <2A0F5F54-E5E3-40CB-B25C-DE93EB26B872@cs.toronto.edu>


On 31-Oct-08, at 1:21 PM, Wu, Kejia wrote:

> Also, can any body tell me whether the random number algorithm in RNG
> package is a pseudorandom one or a real-random one?

You can't generate real-random numbers in software alone. Real random  
number generation relies on sampling some random physical process.  
Google "real random number" and you'll find a number of online sources  
of genuine random numbers, including random.org (which uses  
atmospheric noise) and hotbits (which uses radioactive decay).

> And is there an
> available implementation for Monte Carlo method in NumPy?

Try http://code.google.com/p/pymc/

David