summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* REL: NumPy 1.18.4 release.v1.18.4Charles Harris2020-05-021-1/+1
|
* Merge pull request #16140 from charris/prepare-1.18.4-releaseCharles Harris2020-05-022-32/+46
|\ | | | | REL: Prepare for the NumPy 1.18.4 release.
| * REL: Prepare for the NumPy 1.18.4 release.Charles Harris2020-05-022-32/+46
|/ | | | | - Finish the 1.18.4-notes.rst. - Create the 1.18.4-changelog.rst
* Merge pull request #16132 from charris/backport-16014Charles Harris2020-05-013-19/+141
|\ | | | | Backport 16014
| * Fix other comments from Ross and make it ImportError specificSebastian Berg2020-05-013-7/+7
| |
| * Apply suggestions from code reviewSebastian Berg2020-05-012-7/+6
| | | | | | Co-Authored-By: Ross Barnowski <rossbar@berkeley.edu>
| * DOC: Change import error "howto" to link to troubleshooting docsSebastian Berg2020-05-013-19/+142
|/ | | | | | | | | This makes the C-API failed to load import error link to a new troubleshooting document in the user documentation. It links to the devdocs directly, since it should be possible to update the information at any time (e.g. when a new setup starts making issues). As these are not NumPy issues, their occurance is not tied to the NumPy release process.
* Merge pull request #16109 from seberg/revert-accidental-bool-fixCharles Harris2020-04-301-0/+69
|\ | | | | REV: Reverts side-effect changes to casting
| * REV: Reverts side-effect changes to castingSebastian Berg2020-04-291-0/+69
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This mostly reverts commit 6b954aa47bcce17cd7497d41f818950f847fbd8e to ensure only the longdouble floating point numbers are changed, for which this was a bugfix was relevant. All other numbers should be unchanged (and were faster with the change), but revert since this is a bugfix release. Mainly the string->boolean conversions where accidentally changed to arguably more correct behaviour. But this should not have happend accidentally, especially in a bug-fix branch. Another corner case change is that string->datetime casts within hypothetical C-extensions doing manual C-level casting was fixed to be in line with normal string -> datetime casts (the cast was inconsistently defined). This revert is only applied as a backport. The second change will remain included in the next non-bugfix release while the first is fixed in a more specific in cleaner manner.
* | Merge pull request #16114 from charris/backport-16101Charles Harris2020-04-291-9/+18
|\ \ | | | | | | BLD: put openblas library in local directory on windows
| * | BLD: put openblas library in local directory on windows (#16101)Matti Picus2020-04-291-9/+18
|/ / | | | | | | | | | | | | | | * BLD: put openblas library in local directory on windows * TST: fix broken pip install on windows * BLD: handle 64-bit windows BLAS
* | Merge pull request #16090 from charris/backport-16076Charles Harris2020-04-273-11/+104
|\ \ | | | | | | BUG: random: Generator.integers(2**32) always returned 0.
| * | TST: random: Skip a test if integers are 32 bit.Warren Weckesser2020-04-271-0/+2
| | |
| * | TST: random: Add more repeatability tests for random integers.Warren Weckesser2020-04-272-0/+45
| | | | | | | | | | | | | | | | | | | | | Add repeatability tests for when the range of the integers is `2**32` (and `2**32 +/- 1` for good measure) with broadcasting. The underlying functions called by Generator.integers and random.randint when the inputs are broadcast are different than when the inputs are scalars.
| * | MAINT: random: Add assert() statements.Warren Weckesser2020-04-271-0/+11
| | | | | | | | | | | | | | | Assert that an invalid value (2**n-1 for n = 8, 16, 32, 64) has not been passed to the Lemire function.
| * | BUG: random: Generator.integers(2**32) always returned 0.Warren Weckesser2020-04-272-11/+46
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When the input to Generator.integers was 2**32, the value 2**32-1 was being passed as the `rng` argument to the 32-bit Lemire method, but that method requires `rng` be strictly less then 2**32-1. The fix was to handle 2**32-1 by calling next_uint32 directly. This also works for the legacy code without changing the stream of random integers from `randint`. Closes gh-16066.
* | | Merge pull request #16091 from charris/backport-16077Charles Harris2020-04-271-7/+17
|\ \ \ | |_|/ |/| | BLD: fix path to libgfortran on macOS
| * | BLD: change link, add sha256 checkmattip2020-04-271-2/+8
| | |
| * | BLD: fix path to libgfortran on macOSmattip2020-04-271-6/+10
| |/
* | Merge pull request #16055 from mattip/fix-linux32Sebastian Berg2020-04-231-1/+1
|\ \ | |/ |/| BLD: add i686 for 1.18 builds
| * BLD: add i686 for 1.18 buildsmattip2020-04-231-1/+1
|/
* REL: prepare 1.18.x for further developmentCharles Harris2020-04-194-3/+49
|
* REL: NumPy 1.18.3 release.v1.18.3Charles Harris2020-04-191-1/+1
|
* Merge pull request #16020 from charris/prepare-1.18.3Charles Harris2020-04-194-32/+50
|\ | | | | REL: Prepare for NumPy 1.18.3 release.
| * REL: Prepare for NumPy 1.18.3 release.Charles Harris2020-04-194-32/+50
|/ | | | | | - Create 1.18.3-changelog - Update 1.18.3-release notes. - Clear upcoming_changes
* Merge pull request #16017 from charris/backport-15951Charles Harris2020-04-194-18/+32
|\ | | | | BUG: Alpha parameter must be 1D in `generator.dirichlet`
| * BUG: Alpha parameter must be 1D in `generator.dirichlet` (#15951)panpiort82020-04-194-18/+32
| | | | | | | | | | | | | | | | | | Only one dimensional alpha paramter is currently supported, but higher dimensions were silently allowed and gave an incorrect results. This fixes the regression. In the future, the API could be extended to allow higher dimensional arrays for alpha. Fixes gh-15915
* | Merge pull request #15954 from charris/backport-15876Charles Harris2020-04-114-4/+16
|\ \ | |/ |/| BUG: Check that `pvals` is 1D in `_generator.multinomial`.
| * Add minor tests enhancementPan Jan2020-04-112-2/+2
| |
| * Replace checks with direct function callPan Jan2020-04-112-8/+4
| |
| * Add testing and replace checkPan Jan2020-04-114-4/+16
| |
| * BUG: add check if pvals is 1d array in numpy.random.*.multinomialPan Jan2020-04-112-0/+4
|/
* Merge pull request #15930 from charris/backport-15736Charles Harris2020-04-071-7/+13
|\ | | | | BUG: Guarantee array is in valid state after memory error occurs in getset.c
| * BUG: Guarantee array is in valid state after memory error occurs in getset.cMibu2872020-04-071-7/+13
| | | | | | | | | | | | | | | | | | | | | | | | Currently, in function array_shape_set, ndarray's pointers to dimensions and strides are freed before new array is allocated (Line 71). In case memory error occur, the array is left with dangling pointers. Therefore, we can not recover from such error. * To guarantee array in valid state when memory error occur * Fix: Free cache when nd == 0 to avoid memory leak * Update numpy/core/src/multiarray/getset.c Co-Authored-By: Eric Wieser <wieser.eric@gmail.com>
* | Merge pull request #15929 from charris/backport-15766Charles Harris2020-04-073-81/+30
|\ \ | |/ |/| BUG,MAINT: Remove incorrect special case in string to number casts
| * BUG: Increase default string cast size of longdouble/clongdoubleSebastian Berg2020-04-071-2/+6
| |
| * BUG,MAINT: Remove incorrect special case in string to number castsSebastian Berg2020-04-072-79/+24
|/ | | | | | | | | The string to number casts fall back to using the scalars and the type setitem function to do the cast. However, before calling setitem, they sometimes already called the Python function for string coercion. This is unnecessary. Closes gh-15608
* Merge pull request #15916 from charris/backport-15872Charles Harris2020-04-063-3/+32
|\ | | | | Bug: Fix eigh and cholesky methods of numpy.random.multivariate_normal
| * Bug: Fix eigh mnd cholesky methods of numpy.random.multivariate_normalMax Balandat2020-04-063-3/+32
|/ | | | Fixes #15871
* REL: prepare 1.18.x for further developmentCharles Harris2020-03-174-3/+49
|
* REL: NumPy 1.18.2 release.v1.18.2Charles Harris2020-03-091-1/+1
|
* Merge pull request #15734 from charris/backport-15723Charles Harris2020-03-093-6/+11
|\ | | | | BUG: fix logic error when nm fails on 32-bit
| * BUG: fix logic error when nm fails on 32-bitmattip2020-03-093-6/+11
|/
* Merge pull request #15729 from charris/prepare-1.18.2Charles Harris2020-03-092-32/+43
|\ | | | | REL: Prepare for 1.18.2 release.
| * REL: Prepare for 1.18.2 release.Charles Harris2020-03-092-32/+43
|/ | | | | - Update release notes. - Create changelog
* Merge pull request #15722 from charris/backport-15714Charles Harris2020-03-077-38/+83
|\ | | | | MAINT: use list-based APIs to call subprocesses
| * MAINT: use list-based APIs to call subprocesses (#15714)Sebastian Berg2020-03-077-38/+83
|/ | | | | | | | | * MAINT: use list-based APIs to call subprocesses * TST, MAINT: add a test for mingw32ccompiler.build_import, clean up lib2def Co-authored-by: Matti Picus <matti.picus@gmail.com> Co-authored-by: Eric Wieser <wieser.eric@gmail.com> Also remove `nose` from test_requirements.py.
* Merge pull request #15679 from charris/backport-15553Charles Harris2020-03-011-0/+5
|\ | | | | BUG: Added missing error check in `ndarray.__contains__`
| * BUG: Added missing error check in `ndarray.__contains__` (gh-15553)Ganesh Kathiresan2020-03-011-0/+5
|/ | | | | | Addes missing check for NULL return. resolves #15552 resolves pandas-dev/pandas#31922
* Merge pull request #15676 from charris/backport-15511Charles Harris2020-03-013-92/+72
|\ | | | | MAINT: Large overhead in some random functions
View Lorry Controller Status