summaryrefslogtreecommitdiff
path: root/numpy
Commit message (Collapse)AuthorAgeFilesLines
* BUG: Remove pointer cast from the PyArray_BYTES macro.Frederic2012-11-061-1/+1
| | | | | | | This fixes a compilation problem encountered with numpy 1.7 when the PyArray_BYTES macro was used as an lvalue in Theano. The macro worked as an lvalue in 1.6 and in 1.7 only the struct name was changed, which makes the problem a bit of a mystery, but Theano compiles with this fix.
* Merge pull request #2702 from seberg/issue2700Stefan van der Walt2012-11-062-95/+46
|\ | | | | BUG: Reshape of 0-sized arrays failed to work without copy.
| * MNT: Remove _check_zeros from NewShapeSebastian Berg2012-11-011-92/+36
| | | | | | | | | | _attempt_nocopy_reshape already handles this case fine. The only purpose would be to slightly speed up this special case.
| * BUG: Reshape of 0-sized arrays failed to work without copySebastian Berg2012-11-012-5/+12
| | | | | | | | | | This also adds a check for order=Keeporder which is not supported. "closes Issue #2700"
* | Merge pull request #483 from astrofrog/fix-masked-getitemnjsmith2012-10-312-10/+8
|\ \ | |/ |/| When accessing MaskedArray rows, always return an mvoid object
| * Fixed failing testsThomas Robitaille2012-10-111-4/+4
| |
| * When accessing MaskedArray rows, always return an mvoid objectThomas Robitaille2012-10-111-6/+4
| |
* | Merge pull request #2694 from seberg/cflags2njsmith2012-10-2513-177/+154
|\ \ | | | | | | ENH: Make 1-dimensional axes not matter for contiguous flags
| * | MNT: Remove unnecessary special cases in _UpdateContiguousFlagsSebastian Berg2012-10-231-16/+0
| | | | | | | | | | | | | | | Remove unnecessary special cases for 0 and 1-d arrays as suggested by njsmith.
| * | MNT: Minor cleanups in comments and lowlevel_strided_loops.Sebastian Berg2012-10-223-9/+10
| | |
| * | TST: Check if contiguous flags are correct in various situationsSebastian Berg2012-10-221-0/+32
| | |
| * | BUG: Fix bug with size 1-dims in CreateSortedStridePermSebastian Berg2012-10-215-20/+11
| | | | | | | | | | | | | | | | | | | | | | | | This reverts changes done to CreateSortedStridePerm in commit 9194b3af. The problem is that it would fail for 3x1x3 Fortran order array for example. And special handleing seems unnecessary at least after 1-dim axis not mattering for contiguous flags. This "closes Issue #434"
| * | MNT: Remove unnecessary stride/flags cleanup.Sebastian Berg2012-10-212-40/+5
| | | | | | | | | | | | | | | | | | | | | | | | This code is unnecessary with changed flags behavior. It would only serve the purpose of making strides look nicer for the user. The UpdateFlags was only required since 1-dim axis being removed might change contiguous flags. But this cannot happen now.
| * | API: ctors changed so that contiguous flags ignore 1-dim axisSebastian Berg2012-10-211-10/+30
| | | | | | | | | | | | | | | | | | This changes ctors.c so that new arrays are created in such a way that they are both C- and F-contiguous if possible. Also fixes some corner cases for 0-sized arrays.
| * | API: Change Flags Updateing to allow C-/F-contiguous arraysSebastian Berg2012-10-211-62/+45
| | | | | | | | | | | | | | | This changes UpdateFlags to ignore 1-dimensional axis when setting C-/F-contiguous flags. Updates both flags always now.
| * | TST: Test order='F' combined with C/F-contiguous arrays.Sebastian Berg2012-10-211-0/+10
| | | | | | | | | | | | | | | | | | Both ravel and asarray misbehaved on 1D (or more generally C and F-contiguous arrays if order='F' was specified. Speed test for .tostring seems not feasably.
| * | BUG: Replace ISFORTRAN by IS_F_CONTIGUOUS.Sebastian Berg2012-10-213-6/+6
| | | | | | | | | | | | | | | | | | In a few places ISFORTRAN was used to check for f-contiguouity. This is incorrect, since ISFORTRAN never evaluated to True if the array was also c-contiguous.
| * | API: Change isfortran to f-contig and not c-contigSebastian Berg2012-10-212-5/+6
| | | | | | | | | | | | | | | | | | This was a check for 1-d arrays, this is the generalization to higher dimension if we allow higher dimensioal arrays to be both f- and c-contiguous.
| * | TST: Update ctypes tests for changed contiguous flags setting.Sebastian Berg2012-10-211-2/+2
| | | | | | | | | | | | This is unrelated to what is actually tested here.
| * | ENH: Do not rely on clean strides for loops.Sebastian Berg2012-10-211-28/+18
| |/ | | | | | | | | | | In lowlevel_strided_loops.h, do not rely that the strides for contiguous arrays are set nicely since they can be considered not important if an axis dimensions is zero.
* | Merge pull request #482 from ContinuumIO/void_scalar_hashTravis E. Oliphant2012-10-212-2/+45
|\ \ | |/ |/| BUG: Fix void_scalar hash function to use the elements of the record in ...
| * BUG: Ensure that writeable records are not hashed.Travis E. Oliphant2012-10-132-0/+12
| |
| * TST: Add one more test.Travis E. Oliphant2012-10-091-0/+2
| |
| * BUG: Fix void_scalar hash function to use the elements of the record in the ↵Travis E. Oliphant2012-10-092-2/+31
| | | | | | | | hash computation just like the tuple hash
* | Merge pull request #480 from matthew-brett/fix-distutils-log-warningTravis E. Oliphant2012-10-091-2/+0
|\ \ | | | | | | FIX: remove log line causing warning from disutils
| * | FIX: remove log line causing warning from disutilsMatthew Brett2012-10-061-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Using numpy.distutils through easy_install caused a RuntimeWarning because of a failed import of numpy.distutils. Discussion here: http://thread.gmane.org/gmane.comp.python.numeric.general/51719 The conclusion seemed to be that the safest fix is to remove the one line of logging in the relevant callback. Thanks to Ralf Gommers for the suggestion.
* | | Merge pull request #476 from njsmith/copy-memory-orderTravis E. Oliphant2012-10-093-3/+34
|\ \ \ | | | | | | | | [FIX] preserve memory order in np.copy()
| * | | [FIX] preserve memory order in np.copy()Nathaniel J. Smith2012-10-013-3/+34
| | | | | | | | | | | | | | | | | | | | This switches us back to the behaviour seen in numpy 1.6 and earlier, which it turns out that scikit-learn (and probably others) relied on.
* | | | Merge pull request #466 from seberg/issue465Travis E. Oliphant2012-10-092-3/+18
|\ \ \ \ | | | | | | | | | | BUG: Fill correct strides for ndmin in array creation
| * | | | TST: Test that ndmin honors requested and original order.Sebastian Berg2012-09-291-0/+7
| | | | |
| * | | | BUG: Fill correct strides for ndmin in array creationSebastian Berg2012-09-291-3/+11
| | | | | | | | | | | | | | | | | | | | | | | | | Closes "Issue #465", strides need to be set according to the requested contiguous flags.
* | | | | Merge pull request #454 from seberg/issue380Travis E. Oliphant2012-10-092-4/+10
|\ \ \ \ \ | | | | | | | | | | | | FIX: Issue #380
| * | | | | TST: reshaping of zero strided array (Issue #380)Sebastian Berg2012-09-241-0/+6
| | | | | |
| * | | | | BUG: strides should only be changed if dimensions==1Sebastian Berg2012-09-241-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | This fixes Issue #380".
* | | | | | Merge pull request #444 from 87/fix_array_ifaceTravis E. Oliphant2012-10-093-102/+183
|\ \ \ \ \ \ | |_|_|_|_|/ |/| | | | | Fix for issue #291
| * | | | | TST: Add more tests.Han Genuit2012-09-301-3/+14
| | | | | |
| * | | | | BUG: Disallow shapesize > 1 when coercing scalar to arrayHan Genuit2012-09-301-0/+7
| | | | | |
| * | | | | ENH: Assume object is scalar if no shape and data attribute are givenHan Genuit2012-09-301-19/+24
| | | | | |
| * | | | | BUG: Handle scalars correctly.Han Genuit2012-09-301-2/+2
| | | | | |
| * | | | | BUG: Fixes for Python 3.xHan Genuit2012-09-202-3/+18
| | | | | |
| * | | | | BUG: Make sure empty shapes are disallowed when data is specifiedHan Genuit2012-09-161-0/+10
| | | | | |
| * | | | | TST: Add test for special case in array interfaceHan Genuit2012-09-161-0/+14
| | | | | |
| * | | | | BUG: Make shape check for special case a bit stricterHan Genuit2012-09-161-2/+2
| | | | | |
| * | | | | ENH: Allow objects to use the array interface without specifying dataHan Genuit2012-09-161-105/+124
| | | | | |
* | | | | | BUG: fix in-place build.David Cournapeau2012-10-091-1/+1
| | | | | |
* | | | | | inplace passes testsJohn Salvatier2012-10-082-4/+5
| | | | | |
* | | | | | fixed testsJohn Salvatier2012-10-082-8/+9
| | | | | |
* | | | | | added initial mapiter testsJohn Salvatier2012-10-082-0/+149
| | | | | |
* | | | | | eliminated oned and fancy arguments to MapIterArrayJohn Salvatier2012-10-081-6/+13
| | | | | |
* | | | | | changed error handling for clarityJohn Salvatier2012-10-081-3/+5
| | | | | |
View Lorry Controller Status