On Wed, Dec 2, 2009 at 1:28 PM, Christopher Barker <chris.bar...@noaa.gov>wrote:
> I downloaded rc1, and built it on my PPC OS-X 10.4 box, with Python > 2.5.2 (from python.org). Then ran the tests. I got: > > ---------------------------------------------------------------------- > Ran 2521 tests in 24.804s > > FAILED (KNOWNFAIL=4, SKIP=1, errors=27, failures=2) > <nose.result.TextTestResult run=2521 errors=27 failures=2> > > Many of them look like this: > > ERROR: test_basic (test_defmatrix.TestAlgebra) > ---------------------------------------------------------------------- > Traceback (most recent call last): > File > > "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/numpy/core/tests/test_defmatrix.py", > line 189, in test_basic > mA = matrix(A) > NameError: global name 'matrix' is not defined > > > Some sort of namespace issue? np.matrix does exist. > > David says this is due to a stray old file (see earlier post), you need to clean out the previous numpy installation. > Then there is an issue have no clue about: > > ====================================================================== > FAIL: test_umath.test_nextafterl > ---------------------------------------------------------------------- > Traceback (most recent call last): > File > > "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/nose-0.10.4-py2.5.egg/nose/case.py", > line 182, in runTest > self.test(*self.arg) > File > > "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/numpy/testing/decorators.py", > line 215, in knownfailer > return f(*args, **kwargs) > File > > "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/numpy/core/tests/test_umath.py", > line 866, in test_nextafterl > return _test_nextafter(np.longdouble) > File > > "/Library/Frameworks/Python.framework/Versions/2.5/lib/python2.5/site-packages/numpy/core/tests/test_umath.py", > line 852, in _test_nextafter > assert np.nextafter(one, two) - one == eps > AssertionError > > > Is this something with eps and PPC (endian issue?), maybe? > > It is the odd long double of PPC, also reported earlier. It should probably be marked a known fail on PPC. If David or a developer had a PPC it might could be worked out. OTOH, long double on PPC isn't any sort of IEEE so, strictly speaking, we don't support it. <snip> Chuck
_______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion