summaryrefslogtreecommitdiff
path: root/numpy
Commit message (Collapse)AuthorAgeFilesLines
* Added axis argument to numpy.uniquemartinosorb2016-11-132-116/+259
|
* Merge pull request #8194 from alvarosg/scalar-piecewiseseberg2016-11-092-6/+23
|\ | | | | BUG: np.piecewise not working for scalars
| * BUG: np.piecewise not working for scalarsalvarosg2016-10-262-6/+23
| |
* | Merge pull request #5302 from idfah/masterRalf Gommers2016-11-062-11/+32
|\ \ | | | | | | Fixed meshgrid to return arrays with same dtype as arguments.
| * | Fix to meshgrid allows passing array sub-classesElliott Forney2016-10-101-1/+1
| | |
| * | Fixed meshgrid to return arrays with same dtype as arguments.Elliott M Forney2014-11-202-11/+32
| | |
* | | Merge pull request #6659 from kiwifb/asarrayRalf Gommers2016-11-062-3/+40
|\ \ \ | | | | | | | | BUG: Let linspace accept input that has an array_interface but is not othe…
| * | | Adds a regression test that demonstrates the issue.Erik M. Bray2016-08-291-0/+36
| | | |
| * | | Merge branch 'master' into asarrayFrançois Bissey2016-08-23285-6229/+13835
| |\ \ \
| * | | | Use asanyarray instead of asarrayFrançois Bissey2015-11-111-4/+3
| | | | |
| * | | | Let linspace accept input that has an array_interface but is not otherwise a ↵François Bissey2015-11-101-2/+4
| | | | | | | | | | | | | | | | | | | | numpy or python type.
* | | | | DOC: fix typos in arrayprint docstrings.Ralf Gommers2016-11-061-2/+2
| | | | | | | | | | | | | | | | | | | | Addresses part of issue reported in gh-8224.
* | | | | BUG: Update operator `**` tests for new behavior.Charles Harris2016-11-051-17/+35
| | | | | | | | | | | | | | | | | | | | | | | | | I forgot to keep the test fixes when the scalar `**` operator was refactored the second time.
* | | | | Merge pull request #8240 from charris/prepare-for-1.12-branchCharles Harris2016-11-053-0/+3
|\ \ \ \ \ | | | | | | | | | | | | REL: Prepare for 1.12.x branch
| * | | | | REL: Updates for C-ABI and C-API.Charles Harris2016-11-053-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | No changes in the interfaces but some housekeeping is needed. - Add comments to cversions.txt and setup_common.py - Add NPY_1_12_API_VERSION to numpyconfig.h [ci skip]
* | | | | | Merge pull request #8227 from alimuldal/uniform_inf_checkCharles Harris2016-11-052-0/+6
|\ \ \ \ \ \ | |/ / / / / |/| | | | | BUG: Better check for invalid bounds in np.random.uniform.
| * | | | | BUG: Better check for invalid bounds in np.random.uniform.Alistair Muldal2016-11-022-0/+6
| | | | | | | | | | | | | | | | | | | | | | | | Also check for invalid bounds when low= and high= are arraylike rather than scalar (closes #8226)
* | | | | | Merge pull request #8231 from charris/integer-to-neg-integer-powerCharles Harris2016-11-051-62/+30
|\ \ \ \ \ \ | | | | | | | | | | | | | | ENH: Refactor numpy ** operators for numpy scalar integer powers
| * | | | | | ENH: Refactor numpy ** operators for numpy scalar integer powers.Charles Harris2016-11-011-62/+30
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This change is the upshot of discussion on the numpy mailing list where it was decided that a numpy scalar integer raised to a negative integer power should raise a ValueError, as do Numpy arrays. In particular, we want numpy scalars and numpy 0-d arrays to have similar behavior. Previously zero was returned for this case.
* | | | | | | Merge pull request #8190 from charris/add-fpower-ufuncCharles Harris2016-11-053-1/+80
|\ \ \ \ \ \ \ | | | | | | | | | | | | | | | | ENH: Add a float_power function with at least float64 precision.
| * | | | | | | ENH: Add a float_power function with at least float64 precision.Charles Harris2016-11-023-1/+80
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The new float_power function converts integers and inexact type to inexact types of minimum precision float64 before computing the power. The intent is one should not need to worry about overflow or negative powers when computing a power when type preservation is not a factor. It also implements the popular proposal on the numpy mailing list that the `**` operator should always return at least double precision.
* | | | | | | | Merge pull request #8236 from charris/fix-documentation-prsCharles Harris2016-11-041-28/+33
|\ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | DOC: Fix documentation pull requests.
| * | | | | | | | DOC: Correct documentation of hfft/ihfft.Charles Harris2016-11-041-28/+33
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Corrections to #7492. Correct the examples in the notes specifying the output length for the hfft. Also clean up the documentation a bit. [ci skip]
* | | | | | | | | Merge pull request #8098 from stuartarchibald/wip/np_la_eigh_doc_enhanceCharles Harris2016-11-041-2/+47
|\ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | DOC: Enhance description/usage for np.linalg.eig*h
| * | | | | | | | | DOC: Remove "deprecated" from eigvalsh UPLO.Charles Harris2016-11-041-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The UPLO parameter of eigvalsh was marked as deprecated, apparently in anticipation of convergence with the scipy version which uses a `lower` boolean parameter. There is no deprecation in the code, nor a lower parameter, so remove the deprecation. [ci skip]
| * | | | | | | | | DOC: Enhance description/usage for np.linalg.eig*hStuart Archibald2016-09-291-1/+46
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Proposed documentation enhancement to the `np.linalg.eig*h` routines. The current documentation does not reflect the nature of the underlying LAPACK routines and the way in which they interpet the input. This documentation and examples are with view of adding clarity to what is actually performed.
* | | | | | | | | | Merge pull request #8222 from nouiz/mean_float16Charles Harris2016-11-033-3/+21
|\ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | ENH: Make numpy.mean() do more precise computation
| * | | | | | | | | | DOC: Small fixup to the documentation.Charles Harris2016-11-031-2/+2
| | | | | | | | | | |
| * | | | | | | | | | Keep the subclass and update the docstring per the comment.Frederic Bastien2016-11-032-3/+4
| | | | | | | | | | |
| * | | | | | | | | | Simplify and still reuse out with float16 inputs.Frederic Bastien2016-10-311-4/+5
| | | | | | | | | | |
| * | | | | | | | | | [ENH]Make numpy.mean() do more precise computation without changing the ↵Frederic Bastien2016-10-283-3/+19
| | |_|_|_|/ / / / / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | output dtype that stay in float16.
* | | | | | | | | | DOC: Clarified when a copy is made in numpy.asarray [skip ci]Michael Seifert2016-11-031-2/+2
| |_|/ / / / / / / |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | asarray does copy the input if the specified dtype and/or order differ from the input even if it is an ndarray.
* | | | | | | | | Merge pull request #8200 from skwbc/structure_formatahaldane2016-11-012-49/+66
|\ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | BUG: Fix structured array format functions
| * | | | | | | | | MAINT: Refactor numpy/core/arrayprint.pyShota Kawabuchi2016-11-011-18/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Related to PR #8200
| * | | | | | | | | BUG: Fix array2string for structured array scalarsShota Kawabuchi2016-10-292-15/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | PR #8160 added format function for structured arrays. But it is not applied for structured array scalars. Closes #8172
| * | | | | | | | | BUG: Fix subarray format changed in #8160Shota Kawabuchi2016-10-222-19/+41
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Preserving structured array element format, this commit fixes subarray format changed in PR #8160. This commit also changes iterator for field name from dtype_.descr to dtype_.names (Related to #8174).
* | | | | | | | | | ENH: Deprecation warnings for `/` integer division when running python -3saurabh2016-11-019-63/+145
| |/ / / / / / / / |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When python is invoked with switch -3, it emits waring "classic int division" for strict integer divisions. The same behavior is now implemented to numpy with this fix
* | | | | | | | | Merge pull request #7987 from f0k/advertise-open_memmapCharles Harris2016-10-282-0/+5
|\ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | DOC: See also np.load and np.memmap in np.lib.format.open_memmap
| * | | | | | | | | DOC: Hint at lib.format.open_memmap from np.load and np.memmapJan Schlüter2016-08-292-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | np.load supports memmap_mode to open existing .npy files as memory-mapped files. np.memmap supports creating or loading raw binary files as memory-mapped files. This adds a link to a function for creating memory-mapped files in .npy format.
* | | | | | | | | | Merge pull request #8218 from mattharrigan/ediff1d-performanceCharles Harris2016-10-282-12/+17
|\ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | BUG: ediff1d should return subclasses
| * | | | | | | | | | BUG: return subclasses from ediff1dMattHarrigan2016-10-272-13/+14
| | | | | | | | | | |
| * | | | | | | | | | ENH: fast track default kwargs for ediff1dMattHarrigan2016-10-261-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Special case to_begin and to_end both equal to None, avoiding subsequent steps. Particulary faster for small arrays where overhead plays a big role.
* | | | | | | | | | | DOC: Clean up lomax and genpareto references.wrwrwr2016-10-281-3/+3
| | | | | | | | | | |
* | | | | | | | | | | MAINT, DOC: Update some scipy.stats references.wrwrwr2016-10-281-27/+27
| |_|_|_|_|/ / / / / |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Closes #5813.
* | | | | | | | | | DOC: Patch doc errors for atleast_nd and frombuffergfyoung2016-10-272-4/+4
| |_|_|_|_|_|_|_|/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Closes gh-8214. Closes gh-8215.
* | | | | | | | | Merge pull request #8183 from mattharrigan/ediff1d-performanceStephan Hoyer2016-10-242-14/+33
|\ \ \ \ \ \ \ \ \ | |/ / / / / / / / | | | | | | | | | Ediff1d performance
| * | | | | | | | TST: Added cases for better coverage of ediff1dMattHarrigan2016-10-211-0/+6
| | | | | | | | |
| * | | | | | | | ENH: performance improvement to ediff1dMattHarrigan2016-10-181-14/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Eliminate a copy operation when to_begin or to_end is given. Also use ravel instead of flatiter which is much faster.
* | | | | | | | | Merge pull request #8198 from eric-wieser/warning-on-accumulate-keepdimCharles Harris2016-10-233-9/+34
|\ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | DEP: Deprecate the keepdims argument to accumulate
| * | | | | | | | | DEP: Deprecate the keepdims argument to accumulateEric Wieser2016-10-233-9/+34
| | | | | | | | | |
View Lorry Controller Status