summaryrefslogtreecommitdiff
path: root/numpy/core/SConscript
Commit message (Collapse)AuthorAgeFilesLines
* DEP: Remove scons related files and code.Charles Harris2013-01-131-531/+0
| | | | | | | | | This removes files and code supporting scons builds. After this change numpy will only support builds using distutils or bento. The removal of scons has been discussed on the list several times and a decision has been made that scons support is no longer needed. This was originally discussed for numpy 1.7 and because the distutils and bento methods are still available we are skipping the usual deprecation period.
* ENH: Split sort source file by sort kind.Charles Harris2012-07-111-1/+3
| | | | | | | | | | | | The different sort kinds are now located in quicksort.c.src mergesort.c.src heapsort.c.src Both direct and indirect sorts are included in each file. This cleanup was done so that additional sorts or quickselect could be added without undue clutter.
* BLD: fix Bento and Numscons builds, that were broken in b272bc60.Ralf Gommers2012-06-211-6/+0
|
* Remove PyArray_ReduceWrapper from public APINathaniel J. Smith2012-06-161-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are two reasons to want to keep PyArray_ReduceWrapper out of the public multiarray API: - Its signature is likely to change if/when masked arrays are added - It is essentially a wrapper for array->scalar transformations (*not* just reductions as its name implies -- the whole reason it is in multiarray.so in the first place is to support count_nonzero, which is not actually a reduction!). It provides some nice conveniences (like making it easy to apply such functions to multiple axes simultaneously), but, we already have a general mechanism for writing array->scalar transformations -- generalized ufuncs. We do not want to have two independent, redundant implementations of this functionality, one in multiarray and one in umath! So in the long run we should add these nice features to the generalized ufunc machinery. And in the short run, we shouldn't add it to the public API and commit ourselves to supporting it. However, simply removing it from numpy_api.py is not easy, because this code was used in both multiarray and umath. This commit: - Moves ReduceWrapper and supporting code to umath/, and makes appropriate changes (e.g. renaming it to PyUFunc_ReduceWrapper and cleaning up the header files). - Reverts numpy.count_nonzero to its previous implementation, so that it loses the new axis= and keepdims= arguments. This is unfortunate, but this change isn't so urgent that it's worth tying our APIs in knots forever. (Perhaps in the future it can become a generalized ufunc.)
* ENH: Always include ucsnarray.h and ucsnarray.cCharles Harris2012-04-211-3/+2
| | | | | This removes the include dependency on PY_UNICODE_WIDE in preparation for defining more general unicode functions in the ucsnarray.c file.
* BUG: Fix scons build script so it works with umathmodule.c.Charles Harris2012-02-061-2/+0
|
* ENH: missingdata: Make numpy.all follow the NA || True == True ruleMark Wiebe2011-08-271-0/+4
|
* ENH: missingdata: Rename na_singleton.[ch] to na_object.[ch]Mark Wiebe2011-08-271-1/+1
|
* ENH: missingdata: Move some of the refactored reduction code into the APIMark Wiebe2011-08-271-0/+1
| | | | | | | Doing this so that it can be used both in multiarray and umath, to make writing new reduction operations generally easier. Also made PyArray_Squeeze work with NA-masked arrays.
* ENH: missingdata: Implement routine for array to array assignmentMark Wiebe2011-08-271-0/+1
|
* ENH: missingdata: Implement wheremasked scalar assignment with overwritena=TrueMark Wiebe2011-08-271-0/+1
|
* ENH: missingdata: Writing some low level general array assignment routinesMark Wiebe2011-08-271-0/+1
| | | | | | Some bugs in the NA reductions got tracked back to a need for an assignment operation which doesn't overwrite NA values. Making a decent array assignment API seems like a good aproach to deal with that.
* ENH: missingdata: Add boilerplate for NA singleton typeMark Wiebe2011-08-271-0/+1
|
* ENH: missingdata: Add the NA mask members to PyArrayObjectMark Wiebe2011-08-271-0/+1
|
* STY: Cleanup trailing whitespace.Charles Harris2011-07-071-1/+1
|
* ENH: nditer: Rename nditer.c.src to indicate it has just the templated bitsMark Wiebe2011-07-071-1/+1
|
* ENH: nditer: Move the non-templated API into its own fileMark Wiebe2011-07-071-0/+1
|
* ENH: nditer: Move construction/copy/destruction to its own implementation fileMark Wiebe2011-07-071-0/+1
|
* ENH: umath: Move the type resolution functions into their own fileMark Wiebe2011-07-061-0/+1
|
* ENH: Merge branch 'datetime_dev'Mark Wiebe2011-06-221-0/+1
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Highlights: * Tighten up date unit vs time unit casting rules, and integrate the NPY_CASTING enum deeper into the datetime conversions * Determine a unit when converting from a string array, similar to when converting from lists of strings * Switch local/utc handling to a timezone= parameter, which also accepts a datetime.tzinfo object. This, for example, enables the use of the pytz library with numpy.datetime64 * Remove the events metadata, make the old API functions raise exceptions, and rename the "frequency" metadata name to "timeunit" * Abstract the flexible dtype mechanism into a function, so that it can be more easily changed without having to add code to many places
| * ENH: datetime-autounit: Detect unit from string arrays as wellMark Wiebe2011-06-161-0/+1
| | | | | | | | Also move datetime string functions to their own source file.
* | ENH: Generate the npysort library for linking during build, but do notCharles Harris2011-06-191-8/+1
| | | | | | | | install it.
* | ENH: Remove the _sort module.Charles Harris2011-06-161-11/+1
| |
* | ENH: Make scons build the npysort library.Charles Harris2011-06-161-2/+18
| |
* | ENH: Break out the sort functions into a separate file.Charles Harris2011-06-161-1/+2
| |
* | ENH: Rename the sorting directory to npysort.Charles Harris2011-06-161-1/+1
| |
* | ENH: Rename _sortmodule.c.src and move it into a new sorting directory.Charles Harris2011-06-161-1/+1
|/ | | | Fix the build to deal with the new name and location.
* STY: datetime-feedback: Rename np.busdaydef -> np.busdaycalendarMark Wiebe2011-06-151-1/+1
| | | | | Also rename the busdaydef parameters to busdaycal parameters. This change was motivated by Chuck's code review feedback.
* ENH: datetime-bday: Create the np.busdaydef business day definition objectMark Wiebe2011-06-131-0/+1
|
* ENH: datetime-bday: Add datetime_busday.c/.h, start busday_offset functionMark Wiebe2011-06-101-0/+1
|
* STY: Rename source files for the iterator to nditer_*.*Mark Wiebe2011-03-131-4/+4
|
* BLD: Fix scons and setup.py buildMark Wiebe2011-02-021-1/+6
| | | | | | Moving lowlevel_strided_loops.h to private needed a setup.py update. Added the rest of the needed files in the scons configuration.
* BLD: fix numscons build.David Cournapeau2011-02-021-0/+5
|
* ENH: core: Create half/float16 data typeMark Wiebe2010-12-011-1/+2
|
* BUG: Fix typo in commit r8564-66.rgommers2010-08-011-1/+1
|
* BUG: fix numscons build issue introduced in r8541.rgommers2010-07-311-0/+2
| | | | | The problem was that if endian.h was not detected, @DEFINE_NPY_HAVE_ENDIAN_H@ was still present in _numpyconfig.h, causing compilation to fail.
* BUG: look for endian.h.David Cournapeau2010-07-291-0/+4
| | | | We should use this header in npy_endian.h if available.
* ENH: implement PEP 3118 buffer consumer on Py2.6, adding a simple Memoryview ↵Pauli Virtanen2010-02-201-0/+1
| | | | | | | | | | | | object To use PEP 3118 buffers as ndarray bases, we need to have a way to release the buffers on garbage collection. The easiest way to do this is to use a Memoryview object -- but they are not available on Python 2.6. Hence, we implement a minimal memory view object that only handles the garbage collection, and use it on Python 2.6. It also implements the new buffer protocol, to allow testing the functionality.
* STY: Remove hard tab.Charles Harris2010-02-171-3/+2
|
* BUG: fix numscons build.David Cournapeau2009-12-121-0/+1
|
* BUG: fix numscons build, missing subst in _numpyconfig.hDavid Cournapeau2009-11-291-0/+2
|
* BUG: do not redefine SIZEOF_LONG in config.h if already defined in python ↵David Cournapeau2009-11-271-1/+6
| | | | headers.
* ENH: add a guard to ensure config.h is never included directly.David Cournapeau2009-11-271-0/+6
|
* BUG: (#1221) special case mac os x in numpyconfig.hDavid Cournapeau2009-11-271-2/+2
| | | | | | | | | | Universal builds break the configure stage: we have to harcode the arch-specific values in the case of mac os x, as we have only one configuration stage per compile, but several arch per compile with the braindead -arch machinery. We rename the old numpyconfig.h to a private header, and numpyconfig.h itself post-fix the values in the case of mac os x.
* ENH: make ieee754 a templated file, need for spacing.David Cournapeau2009-11-101-2/+2
|
* ENH: add a define in scons build for long double representation.David Cournapeau2009-11-091-1/+3
|
* ENH: move npy_config to a newly created private header directory.David Cournapeau2009-11-091-1/+1
|
* BUG: do not run long double check for darwin platform (universal build ↵David Cournapeau2009-11-061-1/+2
| | | | breaks it).
* ENH: add long double representation detection to numscons build.David Cournapeau2009-11-061-2/+6
|
* BUG: handle platform without complex support in numscons build.David Cournapeau2009-11-051-0/+4
|