[SciPy-dev] testresults current trunk

[josef.pktd at gmail.com]

On Tue, Nov 17, 2009 at 4:18 PM, Anne Archibald
<peridot.faceted at gmail.com> wrote:
> 2009/11/17  <josef.pktd at gmail.com>:
>> On Tue, Nov 17, 2009 at 4:01 PM, Anne Archibald
>> <peridot.faceted at gmail.com> wrote:
>>> 2009/11/17  <josef.pktd at gmail.com>:
>>>> 2 errors 1 failure: anything new or just a consequence of my numpy
>>>> test failures ?
>>>
>>>> ERROR: test_decomp.test_lapack_misaligned(<function solve at 0x01C55AF0>, (array
>>>> ([[  1.734e-255,   8.189e-217,   4.025e-178,   1.903e-139,   9.344e-101,
>>>> ----------------------------------------------------------------------
>>>> Traceback (most recent call last):
>>>>  File "c:\programs\python25\lib\site-packages\nose-0.11.1-py2.5.egg\nose\case.p
>>>> y", line 183, in runTest
>>>>    self.test(*self.arg)
>>>>  File "C:\Josef\_progs\Subversion\scipy-trunk_after\trunk\dist\scipy-0.8.0.dev6
>>>> 118.win32\Programs\Python25\Lib\site-packages\scipy\linalg\tests\test_decomp.py"
>>>> , line 1106, in check_lapack_misaligned
>>>>    func(*a,**kwargs)
>>>>  File "\Programs\Python25\Lib\site-packages\scipy\linalg\basic.py", line 127, i
>>>> n solve
>>>>  File "\programs\python25\Lib\site-packages\numpy\lib\function_base.py", line 5
>>>> 86, in asarray_chkfinite
>>>>    raise ValueError, "array must not contain infs or NaNs"
>>>> ValueError: array must not contain infs or NaNs
>>>
>>> That's weird - what architecture are you running the tests on? On my
>>> machine I don't get any NaNs or  infs, but the array is made by
>>> reinterpreting a byte array.
>>>
>>> I'm not quite sure what the best way around this is, since the bug is
>>> not always triggered if I supply sensible values.
>>
>> Windows XP notebook, Intel
>
> 64-bit or 32?

32 single core pretty old

>
>> I have several nan errors with the numpy test suite. So this might not
>> be a scipy problem if it doesn't happen on other machines. I never had
>> any problems with linalg and lapack, so it might be something that the
>> test uses.
>
> The test does feed in outrageous values. What LAPACK are you using?

A (very) old ATLAS from the scipy website. It doesn't show the version info,
at least not without a lot of searching.

Josef

>
> Anne
>
>> Josef
>>
>>
>>>
>>> Anne
>>> _______________________________________________
>>> Scipy-dev mailing list
>>> Scipy-dev at scipy.org
>>> http://mail.scipy.org/mailman/listinfo/scipy-dev
>>>
>> _______________________________________________
>> Scipy-dev mailing list
>> Scipy-dev at scipy.org
>> http://mail.scipy.org/mailman/listinfo/scipy-dev
>>
> _______________________________________________
> Scipy-dev mailing list
> Scipy-dev at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-dev
>