[SciPy-Dev] signal.ltisys test crashes on windows

[Warren Weckesser]

josef.pktd at gmail.com wrote:
> On Fri, Jun 4, 2010 at 10:11 AM, Ralf Gommers
> <ralf.gommers at googlemail.com> wrote:
>   
>> The offending test, (0, 3, 3) of:
>>
>> class TestSS2TF:
>>     def tst_matrix_shapes(self, p, q, r):
>>         ss2tf(np.zeros((p, p)),
>>               np.zeros((p, q)),
>>               np.zeros((r, p)),
>>               np.zeros((r, q)), 0)
>>
>>     def test_basic(self):
>>         for p, q, r in [
>>             (3, 3, 3),
>>             (0, 3, 3),
>>             (1, 1, 1)]:
>>             yield self.tst_matrix_shapes, p, q, r
>>
>>
>> The 0 causes an empty array to be passed to ss2tf, which crashes the
>> interpreter on Windows XP for both 2.5 and 2.6. Is the empty array really
>> what was intended with this test?
>>     
>
> replace
>
>   
>>             (0, 3, 3),
>>     
>  by
>   
>>             (1, 3, 3),
>>     
>
>   

Agreed.  0 is a degenerate case.  Perhaps the original author of the 
test expected ss2tf to handle this case cleanly, but it currently doesn't.

Warren

> We had a recent threads about this, and the crash is avoided with
> numpy trunk (raises an exception instead)
>
> But I don't think the empty array is an appropriate test.
>
> Josef
>
>
>   
>> Thanks,
>> Ralf
>>
>>
>> _______________________________________________
>> 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
>