From joel.c.whalen at intel.com  Wed Sep  7 16:05:12 2016
From: joel.c.whalen at intel.com (Whalen, Joel C)
Date: Wed, 7 Sep 2016 20:05:12 +0000
Subject: [SciPy-User] Installing NumPy/SciPy on Windows without the use of
	Pip
Message-ID: <48405334299A4B4A83C81B60118C1A918DD21E@FMSMSX102.amr.corp.intel.com>

Hello,

I hope I am in the right place with my request. I haven't been able to find anything on stack overflow or your website that gives alternatives to building your python package without the use of Pip. My work network does not allow the network traffic of an automated web tool such as Pip, so I've been installing all python packages manually.

The problem is when I go to "python setup.py install" in C:\Python27\Lib\site-packages\numpy-1.11+mkl-cp27-cp27m-win32\numpy, the setup file does not allow you to invoke it directly and instead prints "This is the wrong setup.py file to run". My question is: What *is* the correct setup.py file? I can't seem to find it in the source code or anyone talking about this online. Perhaps I'm approaching this incorrectly, and would appreciate any and all input.

Thank you for your consideration,


-          Joel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20160907/2cd7eb4b/attachment.html>

From jmsachs at gmail.com  Wed Sep  7 16:15:05 2016
From: jmsachs at gmail.com (Jason Sachs)
Date: Wed, 7 Sep 2016 13:15:05 -0700
Subject: [SciPy-User] Installing NumPy/SciPy on Windows without the use
	of Pip
In-Reply-To: <48405334299A4B4A83C81B60118C1A918DD21E@FMSMSX102.amr.corp.intel.com>
References: <48405334299A4B4A83C81B60118C1A918DD21E@FMSMSX102.amr.corp.intel.com>
Message-ID: <CAOo6sOMkJvxr-=NgyvTqGAw+K67UdxD_DP-kphs0ANp4+KXV8g@mail.gmail.com>

Use Anaconda Python! It's much easier to get going with numpy and scipy and
matplotlib.

https://www.continuum.io/anaconda-overview

On Wed, Sep 7, 2016 at 1:05 PM, Whalen, Joel C <joel.c.whalen at intel.com>
wrote:

> Hello,
>
>
>
> I hope I am in the right place with my request. I haven?t been able to
> find anything on stack overflow or your website that gives alternatives to
> building your python package without the use of Pip. My work network does
> not allow the network traffic of an automated web tool such as Pip, so I?ve
> been installing all python packages manually.
>
>
>
> The problem is when I go to ?python setup.py install? in
> C:\Python27\Lib\site-packages\numpy-1.11+mkl-cp27-cp27m-win32\numpy, the
> setup file does not allow you to invoke it directly and instead prints
> ?This is the wrong setup.py file to run?. My question is: What **is** the
> correct setup.py file? I can?t seem to find it in the source code or anyone
> talking about this online. Perhaps I?m approaching this incorrectly, and
> would appreciate any and all input.
>
>
>
> Thank you for your consideration,
>
>
>
> -          Joel
>
> _______________________________________________
> SciPy-User mailing list
> SciPy-User at scipy.org
> https://mail.scipy.org/mailman/listinfo/scipy-user
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20160907/b07f899e/attachment.html>

From matthew.brett at gmail.com  Wed Sep  7 16:22:59 2016
From: matthew.brett at gmail.com (Matthew Brett)
Date: Wed, 7 Sep 2016 13:22:59 -0700
Subject: [SciPy-User] Installing NumPy/SciPy on Windows without the use
	of Pip
In-Reply-To: <CAOo6sOMkJvxr-=NgyvTqGAw+K67UdxD_DP-kphs0ANp4+KXV8g@mail.gmail.com>
References: <48405334299A4B4A83C81B60118C1A918DD21E@FMSMSX102.amr.corp.intel.com>
 <CAOo6sOMkJvxr-=NgyvTqGAw+K67UdxD_DP-kphs0ANp4+KXV8g@mail.gmail.com>
Message-ID: <CAH6Pt5pmH+aVo67wVmFgoA0uq=EeGuvTq_EtT60oHkhJ6aWkcw@mail.gmail.com>

Hi,

On Wed, Sep 7, 2016 at 1:15 PM, Jason Sachs <jmsachs at gmail.com> wrote:
> Use Anaconda Python! It's much easier to get going with numpy and scipy and
> matplotlib.
>
> https://www.continuum.io/anaconda-overview
>
> On Wed, Sep 7, 2016 at 1:05 PM, Whalen, Joel C <joel.c.whalen at intel.com>
> wrote:
>>
>> Hello,
>>
>>
>>
>> I hope I am in the right place with my request. I haven?t been able to
>> find anything on stack overflow or your website that gives alternatives to
>> building your python package without the use of Pip. My work network does
>> not allow the network traffic of an automated web tool such as Pip, so I?ve
>> been installing all python packages manually.
>>
>>
>>
>> The problem is when I go to ?python setup.py install? in
>> C:\Python27\Lib\site-packages\numpy-1.11+mkl-cp27-cp27m-win32\numpy, the
>> setup file does not allow you to invoke it directly and instead prints ?This
>> is the wrong setup.py file to run?. My question is: What *is* the correct
>> setup.py file? I can?t seem to find it in the source code or anyone talking
>> about this online. Perhaps I?m approaching this incorrectly, and would
>> appreciate any and all input.
>>
>>
>>
>> Thank you for your consideration,

>From your path, it looks like you have downloaded one of Christoph
Gohlke's binaries from http://www.lfd.uci.edu/~gohlke/pythonlibs/ and
unzipped it into your site-packages directory.  Is that the case?  If
so, you need to first delete the numpy-1.11+mkl-cp27-cp27m-win32
directory.

Next, I suggest you download all the wheels you need from Christoph's
site to a directory, maybe c:\wheels, then:

cd c:\wheels
pip install numpy-1.11+mkl-cp27-cp27m-win32.whl

etc.

The `setup.py` files are for installing from the source distributions,
but binary distributions like wheels (*.whl) are a better idea on
Windows.

Cheers,

Matthew


From joel.c.whalen at intel.com  Wed Sep  7 16:57:47 2016
From: joel.c.whalen at intel.com (Whalen, Joel C)
Date: Wed, 7 Sep 2016 20:57:47 +0000
Subject: [SciPy-User] Installing NumPy/SciPy on Windows without the
	use	of Pip
In-Reply-To: <CAH6Pt5pmH+aVo67wVmFgoA0uq=EeGuvTq_EtT60oHkhJ6aWkcw@mail.gmail.com>
References: <48405334299A4B4A83C81B60118C1A918DD21E@FMSMSX102.amr.corp.intel.com>
 <CAOo6sOMkJvxr-=NgyvTqGAw+K67UdxD_DP-kphs0ANp4+KXV8g@mail.gmail.com>
 <CAH6Pt5pmH+aVo67wVmFgoA0uq=EeGuvTq_EtT60oHkhJ6aWkcw@mail.gmail.com>
Message-ID: <48405334299A4B4A83C81B60118C1A918DD286@FMSMSX102.amr.corp.intel.com>

Thank you so much! Unfortunately, pip is telling me that the .whl is not a supported wheel on this platform. I'm on 64 bit, and this is true even after I use the amd64.whl. Same error. 

Also deleting that directory in my site-packages folder was a huge pain because Windows didn't even let Administrator approval remove it. I had to rd /s /q to get rid of it.

All in all, windows really doesn't want me to install numpy apparently.

-----Original Message-----
From: SciPy-User [mailto:scipy-user-bounces at scipy.org] On Behalf Of Matthew Brett
Sent: Wednesday, September 7, 2016 1:23 PM
To: SciPy Users List <scipy-user at scipy.org>
Subject: Re: [SciPy-User] Installing NumPy/SciPy on Windows without the use of Pip

Hi,

On Wed, Sep 7, 2016 at 1:15 PM, Jason Sachs <jmsachs at gmail.com> wrote:
> Use Anaconda Python! It's much easier to get going with numpy and 
> scipy and matplotlib.
>
> https://www.continuum.io/anaconda-overview
>
> On Wed, Sep 7, 2016 at 1:05 PM, Whalen, Joel C 
> <joel.c.whalen at intel.com>
> wrote:
>>
>> Hello,
>>
>>
>>
>> I hope I am in the right place with my request. I haven?t been able 
>> to find anything on stack overflow or your website that gives 
>> alternatives to building your python package without the use of Pip. 
>> My work network does not allow the network traffic of an automated 
>> web tool such as Pip, so I?ve been installing all python packages manually.
>>
>>
>>
>> The problem is when I go to ?python setup.py install? in 
>> C:\Python27\Lib\site-packages\numpy-1.11+mkl-cp27-cp27m-win32\numpy, 
>> the setup file does not allow you to invoke it directly and instead 
>> prints ?This is the wrong setup.py file to run?. My question is: What 
>> *is* the correct setup.py file? I can?t seem to find it in the source 
>> code or anyone talking about this online. Perhaps I?m approaching 
>> this incorrectly, and would appreciate any and all input.
>>
>>
>>
>> Thank you for your consideration,

From your path, it looks like you have downloaded one of Christoph Gohlke's binaries from http://www.lfd.uci.edu/~gohlke/pythonlibs/ and unzipped it into your site-packages directory.  Is that the case?  If so, you need to first delete the numpy-1.11+mkl-cp27-cp27m-win32 directory.

Next, I suggest you download all the wheels you need from Christoph's site to a directory, maybe c:\wheels, then:

cd c:\wheels
pip install numpy-1.11+mkl-cp27-cp27m-win32.whl

etc.

The `setup.py` files are for installing from the source distributions, but binary distributions like wheels (*.whl) are a better idea on Windows.

Cheers,

Matthew
_______________________________________________
SciPy-User mailing list
SciPy-User at scipy.org
https://mail.scipy.org/mailman/listinfo/scipy-user

From matthew.brett at gmail.com  Wed Sep  7 17:38:28 2016
From: matthew.brett at gmail.com (Matthew Brett)
Date: Wed, 7 Sep 2016 14:38:28 -0700
Subject: [SciPy-User] Installing NumPy/SciPy on Windows without the use
	of Pip
In-Reply-To: <48405334299A4B4A83C81B60118C1A918DD286@FMSMSX102.amr.corp.intel.com>
References: <48405334299A4B4A83C81B60118C1A918DD21E@FMSMSX102.amr.corp.intel.com>
 <CAOo6sOMkJvxr-=NgyvTqGAw+K67UdxD_DP-kphs0ANp4+KXV8g@mail.gmail.com>
 <CAH6Pt5pmH+aVo67wVmFgoA0uq=EeGuvTq_EtT60oHkhJ6aWkcw@mail.gmail.com>
 <48405334299A4B4A83C81B60118C1A918DD286@FMSMSX102.amr.corp.intel.com>
Message-ID: <CAH6Pt5pyyDoMW7FKhmZMT+Ys8PVEqkx6s187gk9TGypbAeUJSQ@mail.gmail.com>

Hi,

On Wed, Sep 7, 2016 at 1:57 PM, Whalen, Joel C <joel.c.whalen at intel.com> wrote:
> Thank you so much! Unfortunately, pip is telling me that the .whl is not a supported wheel on this platform. I'm on 64 bit, and this is true even after I use the amd64.whl. Same error.
>
> Also deleting that directory in my site-packages folder was a huge pain because Windows didn't even let Administrator approval remove it. I had to rd /s /q to get rid of it.
>
> All in all, windows really doesn't want me to install numpy apparently.

You're nearly there!   First - upgrade pip to the latest version:

python -m pip install --upgrade pip

Then try installing the `*amd64.whl` - if your *Python* is 64-bit.
Remember that your Python may be 32-bit even though your Windows is 64
bit.  Test which, with:

python -c "import platform; print(platform.architecture())"

Cheers,

Matthew


From joel.c.whalen at intel.com  Wed Sep  7 17:44:59 2016
From: joel.c.whalen at intel.com (Whalen, Joel C)
Date: Wed, 7 Sep 2016 21:44:59 +0000
Subject: [SciPy-User] Installing NumPy/SciPy on Windows without the
	use	of Pip
In-Reply-To: <CAH6Pt5pyyDoMW7FKhmZMT+Ys8PVEqkx6s187gk9TGypbAeUJSQ@mail.gmail.com>
References: <48405334299A4B4A83C81B60118C1A918DD21E@FMSMSX102.amr.corp.intel.com>
 <CAOo6sOMkJvxr-=NgyvTqGAw+K67UdxD_DP-kphs0ANp4+KXV8g@mail.gmail.com>
 <CAH6Pt5pmH+aVo67wVmFgoA0uq=EeGuvTq_EtT60oHkhJ6aWkcw@mail.gmail.com>
 <48405334299A4B4A83C81B60118C1A918DD286@FMSMSX102.amr.corp.intel.com>
 <CAH6Pt5pyyDoMW7FKhmZMT+Ys8PVEqkx6s187gk9TGypbAeUJSQ@mail.gmail.com>
Message-ID: <48405334299A4B4A83C81B60118C1A918DD2CB@FMSMSX102.amr.corp.intel.com>

These are some really great tips! Helps me a lot with just learning python in general. 

I don't have internet on this desktop pc due to an internal issue, but I ended up getting numpy installed with an internal tool! It's a cop-out, I know. But I never understood python wheels until you emailed me this, so I appreciate it a lot. 

I also didn't know about that -c argument. That's awesome and quick. 

Thank you for everything,

- Joel

-----Original Message-----
From: SciPy-User [mailto:scipy-user-bounces at scipy.org] On Behalf Of Matthew Brett
Sent: Wednesday, September 7, 2016 2:38 PM
To: SciPy Users List <scipy-user at scipy.org>
Subject: Re: [SciPy-User] Installing NumPy/SciPy on Windows without the use of Pip

Hi,

On Wed, Sep 7, 2016 at 1:57 PM, Whalen, Joel C <joel.c.whalen at intel.com> wrote:
> Thank you so much! Unfortunately, pip is telling me that the .whl is not a supported wheel on this platform. I'm on 64 bit, and this is true even after I use the amd64.whl. Same error.
>
> Also deleting that directory in my site-packages folder was a huge pain because Windows didn't even let Administrator approval remove it. I had to rd /s /q to get rid of it.
>
> All in all, windows really doesn't want me to install numpy apparently.

You're nearly there!   First - upgrade pip to the latest version:

python -m pip install --upgrade pip

Then try installing the `*amd64.whl` - if your *Python* is 64-bit.
Remember that your Python may be 32-bit even though your Windows is 64 bit.  Test which, with:

python -c "import platform; print(platform.architecture())"

Cheers,

Matthew
_______________________________________________
SciPy-User mailing list
SciPy-User at scipy.org
https://mail.scipy.org/mailman/listinfo/scipy-user


From eulergaussriemann at gmail.com  Sat Sep 10 15:28:47 2016
From: eulergaussriemann at gmail.com (David Goldsmith)
Date: Sat, 10 Sep 2016 12:28:47 -0700
Subject: [SciPy-User] Recommendation for optimize.root method to use
Message-ID: <CA+_2Pn3ywHNBVPg1hjD8yppD-69-QOBWkF4Aj7YpOzsUuht6uw@mail.gmail.com>

Hi!  I found
http://docs.scipy.org/doc/scipy/reference/generated/scipy.optimize.root.html#scipy.optimize.root
and thence found the three references therein, which I've downloaded, but
I'm hoping for "instant gratification": my objective function, from R^4 to
R, involves linear combinations of log and atan (neither of these are
composed with the other) with rational functions of linear and quadratic
polynomials as arguments: any recommendation as to which optimize.root
method(s) I should use?  Time is not of the essence, i.e., I don't
necessarily need them to be super-fast.  Thanks!

DLG
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20160910/e0039726/attachment.html>

From eulergaussriemann at gmail.com  Sun Sep 11 14:52:43 2016
From: eulergaussriemann at gmail.com (David Goldsmith)
Date: Sun, 11 Sep 2016 11:52:43 -0700
Subject: [SciPy-User] Recommendation for optimize.root method to use
In-Reply-To: <CA+_2Pn3ywHNBVPg1hjD8yppD-69-QOBWkF4Aj7YpOzsUuht6uw@mail.gmail.com>
References: <CA+_2Pn3ywHNBVPg1hjD8yppD-69-QOBWkF4Aj7YpOzsUuht6uw@mail.gmail.com>
Message-ID: <CA+_2Pn0YxOZ9BZtkAK3jN42BEETuFOwc6SaUA=UCStN30RDj1Q@mail.gmail.com>

Did I embarrass myself (again) by posting under-informed: I want to solve
an over-determined highly-transcendental system (arising from a non-linear
least-squares minimization) and according to Wikipedia, the standard
(non-guess-and-check) way is to first linearize and then solve that system
using standard approaches, then shift a little, re-solve and check, repeat,
so I guess I don't need the non-linear solver after all...though it sounds
like I might want to use a "fancy" Jacobian estimator: any recommendations
for that?  Thanks!

DLG

On Saturday, September 10, 2016, David Goldsmith <
eulergaussriemann at gmail.com> wrote:

> Hi!  I found http://docs.scipy.org/doc/scipy/reference/generated/
> scipy.optimize.root.html#scipy.optimize.root and thence found the three
> references therein, which I've downloaded, but I'm hoping for "instant
> gratification": my objective function, from R^4 to R, involves linear
> combinations of log and atan (neither of these are composed with the other)
> with rational functions of linear and quadratic polynomials as arguments:
> any recommendation as to which optimize.root method(s) I should use?  Time
> is not of the essence, i.e., I don't necessarily need them to be
> super-fast.  Thanks!
>
> DLG
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20160911/c7edd2fd/attachment.html>

From yellowhat46 at gmail.com  Sat Sep 24 07:01:52 2016
From: yellowhat46 at gmail.com (Vasco Gervasi)
Date: Sat, 24 Sep 2016 13:01:52 +0200
Subject: [SciPy-User] f2py and ifort
Message-ID: <CANtTwcFicGj-CG43YFZWVKyXHr9eP=wrjHSkMN1-6+X3cs8UWw@mail.gmail.com>

Hi all,
I am trying to compile a python module using f2py and ifort.
Attached you will found a simple fortran subroutine to use as test.

This is the command that I use to compile the module:

> python C:\Anaconda2\Scripts\f2py.py -c -m foo foo.f90 --fcompiler=intelvem


The module works on the same machine that I use to compile the module, but
on another machine with just anaconda and mingw (without visual studio and
ifort) it returns:

> DLL load failed


Even if I try to add the option --opts="-static" it returns the same error.

Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20160924/727c886c/attachment.html>
-------------- next part --------------
      subroutine asd(a, b)

      implicit none
      real*8,  intent(in) :: a
      real*8, intent(out) :: b

      b = a + 1

      return
      end subroutine asd

From ccordoba12 at gmail.com  Tue Sep 27 04:37:04 2016
From: ccordoba12 at gmail.com (=?UTF-8?Q?Carlos_C=c3=b3rdoba?=)
Date: Tue, 27 Sep 2016 03:37:04 -0500
Subject: [SciPy-User] [ANN] Spyder 3.0 is released!
Message-ID: <d159e537-61fc-bf12-dc18-f1446d7277b2@gmail.com>

Hi all,

On the behalf of the Spyder Project Contributors 
(https://github.com/spyder-ide/spyder/graphs/contributors), I?m pleased 
to announce that Spyder 3.0 has been released and is available for 
Windows, GNU/Linux and MacOS X: 
https://github.com/spyder-ide/spyder/releases

Spyder is a free, open-source (MIT license) interactive development 
environment for the Python language with advanced editing, interactive 
testing, debugging and introspection features. It was designed to 
provide MATLAB-like features integrated help, interactive console, 
variable explorer with GUI-based editors for NumPy arrays and Pandas 
dataframes), it is strongly oriented towards scientific computing and 
software development.

This release represents more than two years of development since version 
2.3.0 was released, and it introduces major enhancements and new 
features. The most important ones are:

  * Third-party plugins: External developers can now create plugins that
    extend Spyder in novel and interesting ways. For example, we already
    have plugins for the line-profiler and memory-profiler projects, and
    also a graphical frontend for the conda package manager. These
    plugins can be distributed as pip and/or conda packages for authors
    convenience.
  * Improved projects support: Projects have been revamped and improved
    significantly in Spyder 3.0. With our new projects support, people
    will have the possibility of easily working on different coding
    efforts at the same time. That?s because projects save the state of
    open files in the Editor and allow Python packages created as part
    of the project to be imported in our consoles.
  * Support for much more programming languages: Spyder relies now on
    the excellent Pygments library to provide syntax highlight and
    suggest code completions in the Editor, for all programming
    languages supported by it.
  * A new file switcher: Spyder 3.0 comes with a fancy file switcher,
    very similar in spirit to the one present in Sublime Text. This is a
    dialog to select among the open files in the Editor, by doing a
    fuzzy search through their names. It also lets users to view the
    list of classes, methods and functions defined in the current file,
    and select one of them. This dialog is activated with |Ctrl+P|.
  * A Numpy array graphical builder: Users who need to create NumPy
    arrays in Spyder for matrices and vectors can do it now in a
    graphical way by pressing |Ctrl+M| in the Editor or the Consoles.
    This will open an empty 2D table widget to be filled with the data
    required by the user.
  * A new icon theme based on FontAwesome.
  * A new set of default pane layouts for those coming from Rstudio or
    Matlab (under |View > Window layouts|).
  * A simpler and more intuitive way to introduce keyboard shortcuts.
  * Support for PyQt5, which fixes problems in MacOS X and in high
    definition screens.

For a complete list of changes, please see here:

https://github.com/spyder-ide/spyder/blob/3.x/CHANGELOG.md

Spyder 2.3 has been a huge success (being downloaded almost 550,000 
times!) and we hope 3.0 will be as successful as it. For that we fixed 
203 important bugs, merged 218 pull requests from about 40 authors and 
added almost 2850 commits between these two releases.

Don?t forget to follow Spyder updates/news on the project Github 
website: https://github.com/spyder-ide/spyder

Last, but not least, we welcome any contribution that helps making 
Spyder an efficient scientific development/computing environment. Join 
us to help creating your favorite environment!

Enjoy!
-Carlos

?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20160927/33651e33/attachment.html>

From cimrman3 at ntc.zcu.cz  Fri Sep 30 03:50:31 2016
From: cimrman3 at ntc.zcu.cz (Robert Cimrman)
Date: Fri, 30 Sep 2016 09:50:31 +0200
Subject: [SciPy-User] ANN: SfePy 2016.3
Message-ID: <a6188747-f3b3-5e75-c8b9-d8c3f98d0178@ntc.zcu.cz>

I am pleased to announce release 2016.3 of SfePy.

Description
-----------

SfePy (simple finite elements in Python) is a software for solving systems of
coupled partial differential equations by the finite element method or by the
isogeometric analysis (limited support). It is distributed under the new BSD
license.

Home page: http://sfepy.org
Mailing list: http://groups.google.com/group/sfepy-devel
Git (source) repository, issue tracker: http://github.com/sfepy/sfepy

Highlights of this release
--------------------------

- Python 3 support
- testing with Travis CI
- new classes for homogenized coefficients
- using argparse instead of optparse

For full release notes see http://docs.sfepy.org/doc/release_notes.html#id1
(rather long and technical).

Cheers,
Robert Cimrman

---

Contributors to this release in alphabetical order:

Robert Cimrman
Jan Heczko
Thomas Kluyver
Vladimir Lukes