summaryrefslogtreecommitdiff
path: root/numpy/testing/utils.py
Commit message (Collapse)AuthorAgeFilesLines
* ENH: testing: add assert_tol_equal for testing array equality with specified ↵Pauli Virtanen2010-07-281-1/+37
| | | | tolerances
* BUG: Ignore "invalid value" from abs in testing/utils.pyCharles Harris2010-02-211-2/+7
|
* BUG: More workarounds for np.isinf warning in tests.Charles Harris2010-02-211-8/+16
|
* more docstring updates from pydoc website (thanks to everyone who contributed!)Jarrod Millman2010-02-171-17/+99
|
* ENH: handle complex input for assert_array_almost_equal_nulp.David Cournapeau2010-02-091-3/+6
|
* BUG: fix typo.David Cournapeau2010-02-091-1/+1
|
* fixed a whole bunch of doctestsPaul Ivanov2009-12-281-0/+1
|
* BUG: make assert_equal and assert_almost_equal always display err_msgPauli Virtanen2009-12-071-8/+5
|
* Fix tests now that ufuncs raise NotImplementedError.Travis Oliphant2009-12-041-6/+4
|
* ENH: add an assert_warns testing utility.David Cournapeau2009-11-231-1/+25
|
* REF: move warning context manager into utils.David Cournapeau2009-11-231-0/+85
|
* BUG: use %g format in assert_array_*_nulp (fix #1297, thanks to Neil Muller)Pauli Virtanen2009-11-131-2/+2
|
* ENH: remove any mention of original python spacing, use ufunc everywhere.David Cournapeau2009-11-101-33/+2
|
* BUG: wrong exception raised in spacing for incompatible dtype.David Cournapeau2009-10-301-2/+0
|
* STY: remove whitespace.David Cournapeau2009-10-301-3/+3
|
* ENH: add assert_array_max_ulp comparison function.David Cournapeau2009-10-301-1/+12
| | | | | This new comparison raises an error if the number of representable numbers between two arrays exceeds a tolerance.
* ENH: add robust comparison function for floating numbers.David Cournapeau2009-10-301-1/+80
| | | | | | assert_array_almost_equal_nulp use spacing so that a single tolerance number can be used independently on the amplitude of the floating point number.
* ENH: add numpy implementation of F90 spacing function.David Cournapeau2009-10-301-0/+33
|
* ENH: add a function to compute the signed-magnitude interpretation of the ↵David Cournapeau2009-10-301-0/+24
| | | | binary repr of a float.
* Don't include assert_valid_refcount in numpy.testing.*Pauli Virtanen2009-10-101-2/+6
| | | | It's a private function used only in two internal regression tests.
* Docstring update: testingPauli Virtanen2009-10-021-13/+128
|
* BUG: fix list/tuple handling in test_almost_equal.David Cournapeau2009-09-211-1/+2
| | | | We now forward tuple/list instances to test_array_almost_equal.
* All non core regressions tests moved to their respective modules.David Cournapeau2009-09-161-1/+16
|
* Import issubdtype from numpy.core directly so that testing.utils does not ↵David Cournapeau2009-09-161-3/+4
| | | | depend on numpy.lib.
* Hmm, somehow a reference to math.fabs slipped by and didn't show here. TryCharles Harris2009-08-031-1/+1
| | | | again.
* Try another fix for BSD problem. This one uses numpy functions, which isCharles Harris2009-08-031-4/+3
| | | | probably not the best...
* Test fix for BSD buildbot error.Charles Harris2009-08-031-1/+4
|
* Fix small bug in assert_almost_equal.Charles Harris2009-08-031-2/+2
| | | | | | | The comparison of complex numbers is still not quite right because errors can propagate between the real and imaginary parts. They can't be checked separately in general. However, this fix gets rid of the test errors currently showing.
* BUG: handle inf/nan correctly in assert_array_almost_equal.David Cournapeau2009-07-281-1/+31
|
* BUG: fix nan/inf handling for complex dtypes.David Cournapeau2009-07-281-0/+31
|
* Handle complex special values and negative zero.David Cournapeau2009-07-271-4/+43
| | | | | Complex with nan/inf are correctly handled in assert_equal, as well as negative zero.
* Handle nan and inf in assert_equal.David Cournapeau2009-07-271-1/+24
|
* Fix header for assert_array_almost_equal.David Cournapeau2009-07-271-2/+2
|
* BUG: handle nan/inf in assert_approx_equal.David Cournapeau2009-07-271-0/+14
|
* BUG: assert_array_compare did not raise an exception when the nan indexes of ↵David Cournapeau2009-07-271-0/+1
| | | | both arguments did not match.
* Handle nan and inf in assert_almost_equal.David Cournapeau2009-07-271-0/+49
|
* Make testing.rundocs to raise error on failures; otherwise Nose hides themPauli Virtanen2009-04-241-9/+18
| | | | Also drop Python < 2.4 compatibility.
* Fix spelling.Charles Harris2009-04-091-1/+1
|
* Merge from the doc wikiPauli Virtanen2009-03-241-9/+300
|
* Give assert_ a default message of ''.Charles Harris2009-03-141-1/+1
|
* Fixed #745: make assert_array_almost_equal work with object arraysPauli Virtanen2009-03-091-2/+6
|
* Add assert_ as a release safe version of assert for running tests.Charles Harris2009-03-091-1/+8
|
* Allow subclasses of arrays in testing.Stefan van der Walt2009-02-221-3/+3
|
* Put measure in numpy.testing namespace.David Cournapeau2008-10-281-1/+1
|
* Removed unused imports.Alan McIntyre2008-09-091-1/+0
|
* Make testing functions work when python is called with the -OO flag.Charles Harris2008-08-291-15/+27
|
* ran reindentJarrod Millman2008-08-081-1/+1
|
* Merge from documentation editor.Stefan van der Walt2008-08-051-2/+34
|
* Delay import from inspect to reduce startup time.Alan McIntyre2008-07-301-1/+4
|
* Delay import of difflib to reduce startup time.Alan McIntyre2008-07-301-1/+3
|