summaryrefslogtreecommitdiff
path: root/numpy/lib
Commit message (Collapse)AuthorAgeFilesLines
* Unique1d will now return unique as well as reverse indices. Fix order ofStefan van der Walt2008-08-132-12/+53
| | | | returns [patch by Robert Cimrman].
* Disable memmap test on windows because it crashes the testsuite. This has to ↵David Cournapeau2008-08-131-31/+34
| | | | be fixed before a rc of numpy 1.2.0, though (see#827).
* Add Roberto de Almeida's Arrayterator.Stefan van der Walt2008-08-133-0/+190
|
* More consistent nan-operations.Stefan van der Walt2008-08-122-20/+38
|
* breaking API for medianJarrod Millman2008-08-081-12/+9
|
* ran reindentJarrod Millman2008-08-0815-70/+61
|
* Remove piece-by-piece docstring assembly; the full texts were inserted in ↵Pauli Virtanen2008-08-081-79/+0
| | | | the docstrings in r5610
* added comment in histogram warning.dhuard2008-08-051-0/+2
|
* Follow-up on changes to histogram semantics.dhuard2008-08-052-59/+85
| | | | `new` is now set to None by default, which triggers the new behaviour and prints a warning.
* Merge from documentation editor.Stefan van der Walt2008-08-0516-1044/+3857
|
* Defer tempfile import to improve startup time.Stefan van der Walt2008-07-311-1/+1
|
* Fix string type comparisons.Stefan van der Walt2008-07-301-4/+4
|
* Added tests to improve coverage.Alan McIntyre2008-07-241-25/+44
| | | | Converted tests from doctests to unit tests.
* Standardized NumPy import as "import numpy as np".Alan McIntyre2008-07-232-84/+87
| | | | Moved unique1d benchmarking code to new benchmarks directory.
* Fix doctest command continuation line.Alan McIntyre2008-07-231-1/+1
|
* Added tests to improve coverage of numpy.lib.Alan McIntyre2008-07-223-1/+60
|
* Clean up loadtxt. Fix Python 2.3 compatibility.Stefan van der Walt2008-07-221-9/+10
|
* Apply Stefan's patch for Ryan's loadtext fix.Charles Harris2008-07-222-34/+87
|
* Committed patch from Ryan May. It fixes error in loadtxt occurring when ↵dhuard2008-07-222-2/+17
| | | | | | usecols is not None and dtypes are given. I added the test suggested by Ryan.
* Removed unused private function _asarray1d.Alan McIntyre2008-07-211-8/+0
|
* Add a pydoc import in lookfor; missed in the import refactoring.Pauli Virtanen2008-07-201-0/+2
|
* BUG: Allow longs as well as ints to satisfy Win64.Robert Kern2008-07-201-1/+1
|
* Make use of assert_raises from numpy.testing, and added run_module_suite to ↵Alan McIntyre2008-07-141-3/+5
| | | | | | support running test module stand-alone.
* Added tests to improve coverage.Alan McIntyre2008-07-135-4/+254
| | | | Renamed TestPiecewise methods so that they will be picked up by nose.
* Fix doctests to run properly under new execution context.Alan McIntyre2008-07-131-4/+4
|
* Make doctests pass under new execution context.Alan McIntyre2008-07-092-7/+7
|
* Use TestCase instead of NumpyTestCase.Stefan van der Walt2008-07-091-1/+1
|
* Simplify a bit more scons scripts.David Cournapeau2008-07-091-2/+1
|
* Use NumpyPythonExtension instead of DistutilsPythonExtension where possible.David Cournapeau2008-07-091-4/+2
|
* Piecewise should not expose raw memory. Closes #798.Stefan van der Walt2008-07-082-11/+54
|
* Removed convdtype.py as discussed on numpy-discussion list (code was ↵Alan McIntyre2008-07-051-65/+0
| | | | | | intended for the Numeric to numpy transition, but was never used).
* Use the implicit "import numpy as np" made available to all doctests instead Alan McIntyre2008-07-057-99/+93
| | | | | of explicit imports or dependency on the local scope where the doctest is defined..
* BUG: Correctly stub out urllib2.urlopen() for tests given the refactoring of ↵Robert Kern2008-07-032-4/+10
| | | | the local imports.
* Clean up test output such that a completely-passing test suite has no ↵Robert Kern2008-07-031-2/+7
| | | | extraneous output.
* BUG: need to create exceptions correctly.Robert Kern2008-07-032-11/+17
|
* ENH: Add broadcast_arrays() function to expose broadcasting to pure Python ↵Robert Kern2008-07-033-0/+317
| | | | functions that cannot be made to be ufuncs.
* Added note to numpy developers about the policy on using finfo in numpy's ↵Robert Kern2008-07-031-0/+7
| | | | codebase.
* Reduce numpy's import times by delaying a few time consuming imports to the ↵Robert Kern2008-07-035-19/+26
| | | | point of actual use and global instantiations of finfo. Thanks to David Cournapeau for tracking down and fixing the import part of the problem.
* Remove uses of set_package_path, set_local_path, restore_path.Alan McIntyre2008-07-039-103/+53
| | | | | Clean up and (somewhat) standardize test module imports. Removed unneeded reload calls.
* Use the implicit "import numpy as np" made available to all doctests instead ↵Alan McIntyre2008-07-035-48/+39
| | | | | | | of an explicit import. Remove ">>>" from bartlett plotting example since it currently requires matplotlib.
* Use join instead of += to build string.Charles Harris2008-06-291-3/+4
|
* Shorten long lines.Charles Harris2008-06-281-39/+58
|
* Fix ticket #828 by explicitly sorting keys instead of relying on pprint.Charles Harris2008-06-281-8/+12
| | | | Thanks to Neil Muller for the analysis and patch.
* Restore old test framework classes.Alan McIntyre2008-06-2116-19/+24
| | | | | | | | | | | Added numpy.testing.run_module_suite to simplify "if __name__ == '__main__'" boilerplate code in test modules. Removed numpy/testing/pkgtester.py since it just consisted of an import statement after porting SciPy r4424. Allow numpy.*.test() to accept the old keyword arguments (but issue a deprecation warning when old arguments are seen). numpy.*.test() returns a test result object as before. Fixed typo in distutils doc.
* Fix piecewise to handle 0-d inputs.Travis Oliphant2008-06-172-1/+27
|
* Switched to use nose to run tests. Added test and bench functions to all ↵Alan McIntyre2008-06-1716-302/+356
| | | | modules.
* Adapat numpy.lib scons build to new build_dir conventions.David Cournapeau2008-06-121-2/+2
|
* Adapt numpy.lib to new scons build_dir behavior.David Cournapeau2008-06-122-9/+11
|
* Merged revisions 5204-5257 via svnmerge from David Cournapeau2008-06-071-4/+1
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | http://svn.scipy.org/svn/numpy/branches/cdavid ........ r5205 | cdavid | 2008-05-20 17:14:30 +0900 (Tue, 20 May 2008) | 3 lines Initialized merge tracking via "svnmerge" with revisions "1-5204" from http://svn.scipy.org/svn/numpy/trunk ........ r5206 | cdavid | 2008-05-20 17:17:27 +0900 (Tue, 20 May 2008) | 7 lines Current handling of bootstrapping is flawed: I should handle it at the distutils level, not at the scons level. This is the first step to detect bootstrapping at distutils level, and pass its state to scons through command line. ........ r5207 | cdavid | 2008-05-20 17:35:01 +0900 (Tue, 20 May 2008) | 1 line Fix typo when passing bootstrapping option to scons. ........ r5208 | cdavid | 2008-05-20 17:41:11 +0900 (Tue, 20 May 2008) | 5 lines Do not mess with __NUMPY_SETUP__ in scons scripts anymore: this is handled in numscons. ........ r5209 | cdavid | 2008-05-20 17:43:46 +0900 (Tue, 20 May 2008) | 1 line Forgot one file in lapack_lite when no LAPACK is available. ........ r5210 | cdavid | 2008-05-20 18:24:38 +0900 (Tue, 20 May 2008) | 1 line Handle fortran compiler on open-solaris ........
| * Do not mess with __NUMPY_SETUP__ in scons scripts anymore: this is handled inDavid Cournapeau2008-05-201-4/+1
| | | | | | | | | | numscons.