[SciPy-Dev] correlate complex192 failures on win32

[Ralf Gommers]

Hi Warren,

In r6982 you added correlate/convolve tests that are now failing on Windows
for longdouble. Can you please have a look at that? If you don't have time
soon I think it can be marked knownfail for 0.9.0 on Windows.

Offending code:
+
+# Create three classes, one for each complex data type:
TestCorrelateComplex64,
+# TestCorrelateComplex128 and TestCorrelateComplex256.
+# The second number in the pairs is used in the 'decimal' keyword argument
of
+# the  array comparisons in the tests.
+for i, decimal in [(np.csingle, 5), (np.cdouble, 10), (np.clongdouble,
15)]:


Resulting in 16 failures like:

======================================================================
FAIL: test_rank1_full (test_signaltools.TestCorrelateComplex192)
----------------------------------------------------------------------
Traceback (most recent call last):
  File
"Z:\Users\rgommers\.wine\drive_c\Python26\lib\site-packages\scipy\signal\tests\test_signaltools.py",
line 630, in test_rank1_full
    assert_array_almost_equal(y, y_r, decimal=self.decimal)
  File
"Z:\Users\rgommers\.wine\drive_c\Python26\lib\site-packages\numpy\testing\utils.py",
line 774, in assert_array_almost_equal
    header='Arrays are not almost equal')
  File
"Z:\Users\rgommers\.wine\drive_c\Python26\lib\site-packages\numpy\testing\utils.py",
line 618, in assert_array_compare
    raise AssertionError(msg)
AssertionError:
Arrays are not almost equal

(mismatch 11.7647058824%)
 x: array([0.17161931+0.20853512j, 0.22923514+0.78246389j,
       -1.3237712+0.47060531j, -3.2990924+-0.50721318j,
       -2.9089895+0.084727311j, 1.5646499+-0.56960514j,...
 y: array([0.17161931+0.20853512j, 0.22923514+0.78246389j,
       -1.3237712+0.47060531j, -3.2990924+-0.50721318j,
       -2.9089895+0.084727311j, 1.5646499+-0.56960514j,...

Ralf
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/scipy-dev/attachments/20110116/3054c9bc/attachment.html>