summaryrefslogtreecommitdiff
path: root/numpy
Commit message (Collapse)AuthorAgeFilesLines
* DOC: minor documentation fixesPauli Virtanen2011-06-253-3/+5
|
* BUG: deal with shared lib extension in a single place. Fix ctypes.load_library.rgommers2011-06-235-31/+62
| | | | | This is related to PEP 3149, tags in shared library extensions. Only applies to Linux (for now). See also #1749.
* Merge branch 'struct_cleanup'Mark Wiebe2011-06-2315-178/+955
|\
| * ENH: dtype-struct: Tweaks based on Chucks feedbackMark Wiebe2011-06-232-63/+51
| | | | | | | | | | Now the dict repr mode for struct dtypes always includes the itemsize= parameter, so that it will always produce the exact same memory layout.
| * DOC: dtype-struct: Document the changes to struct dtypesMark Wiebe2011-06-222-2/+11
| |
| * ENH: dtype-struct: Some tweaks based on Chuck's feedbackMark Wiebe2011-06-221-18/+23
| |
| * ENH: dtype-struct: Move the 'itemsize' from a parameter into the struct ↵Mark Wiebe2011-06-222-84/+86
| | | | | | | | descriptor
| * ENH: dtype-struct: Ensure alignment of nested struct dtypesMark Wiebe2011-06-222-5/+40
| |
| * ENH: dtype-struct: Add validation that object fields have no overlapMark Wiebe2011-06-222-4/+111
| |
| * ENH: dtype: Allow unions and out-of-order fieldsMark Wiebe2011-06-212-25/+68
| | | | | | | | | | This will require validation that object dtypes don't overlap with other fields in a follow-on commit.
| * BUG: dtype: Struct dtype size wasn't being padded for alignment (Ticket #1790)Mark Wiebe2011-06-212-10/+50
| | | | | | | | | | | | In the case referenced in ticket #1790, the itemsize was set before the padded size was calculated, so the intent was clearly that it should have been aligned.
| * ENH: dtype: Add an itemsize= keyword parameter to the dtype constructorMark Wiebe2011-06-213-44/+144
| | | | | | | | | | The 'struct'-based repr adds this parameter, as the general way to be able to reconstruct the dtype.
| * BUG: dtype: Fix up the list-based struct dtype string representationMark Wiebe2011-06-215-21/+56
| |
| * ENH: dtype: Fix repr of structured dtypes (Fixes #1619)Mark Wiebe2011-06-213-73/+440
| | | | | | | | | | | | | | This includes preserving the 'align=True' parameter through the new NPY_ALIGNED_STRUCT dtype flag. The code now detects whether the list version of structured dtype construction would match the data layout, to choose between list and struct based construction.
| * ENH: dtype: Use PyDataType_HASFIELDS, add a flag constant for aligned structMark Wiebe2011-06-218-35/+81
| |
* | Merge branch 'tighten_casting'Mark Wiebe2011-06-237-29/+88
|\ \
| * | TST: datetime: Fix the tests to respect the date unit/time unit boundaryMark Wiebe2011-06-221-15/+24
| | |
| * | ENH: ma: Fix up ma and its tests to work with default same_kind castingMark Wiebe2011-06-222-4/+5
| | |
| * | ENH: core: Change the ones_like ufunc so it doesn't do type checkingMark Wiebe2011-06-223-1/+31
| | | | | | | | | | | | | | | Since ones_like is a ufunc (unlike zeros_like and empty_like), it by default checks the casting rule. This disables that check completely.
| * | BUG: core: Make slow clip use casting='unsafe' for compatibilityMark Wiebe2011-06-221-1/+24
| | |
| * | ENH: core: Change the default casting rule from 'unsafe' to 'same_kind'Mark Wiebe2011-06-221-8/+4
| | | | | | | | | | | | | | | This is an experimental change, to see what kind of ripple effects tightening up the semantics of NumPy just a little bit produce.
* | | BUG: datetime-print: str() and repr() weren't handling the local tz properlyMark Wiebe2011-06-222-6/+32
|/ /
* | ENH: Merge branch 'datetime_dev'Mark Wiebe2011-06-2227-2408/+3555
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 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: Applying changes for Chuck and Derek's feedbackMark Wiebe2011-06-212-22/+45
| | |
| * | ENH: datetime-strings: Support casting to/from unicode arraysMark Wiebe2011-06-208-43/+213
| | |
| * | ENH: dtype-loops: Move the dtype copy/swap function factory into a functionMark Wiebe2011-06-204-72/+107
| | |
| * | BUG: datetime: Mac OS X doesn't support strnlen, working around thatMark Wiebe2011-06-202-12/+15
| | |
| * | ENH: convert-dtype: Abstract the flexible dtype mechanism into a functionMark Wiebe2011-06-1710-259/+571
| | | | | | | | | | | | | | | | | | | | | | | | | | | Since the datetime adds a new flexible dtype (datetime with generic units), updating all the places where flexible dtypes are adjusted was error-prone. Thus, this has been moved to a single place. At the same time, I've added sizes for the various number types, so they don't produce size-one strings by default anymore
| * | ENH: datetime-meta: Rename the metadata from frequency to timeunitMark Wiebe2011-06-171-1/+1
| | |
| * | DEP: datetime-api: Make the old API functions in datetime.c raise exceptionsMark Wiebe2011-06-171-232/+16
| | |
| * | ENH: datetime-casting: Add casting parameter to 'datetime_as_string'Mark Wiebe2011-06-175-45/+150
| | | | | | | | | | | | | | | | | | The default is 'same_kind', which allows data loss, but prevents crossing the date unit/time unit boundary. As a special case, printing dates with a timezone specified requires 'unsafe' casting.
| * | ENH: datetime-meta: Remove 'events' from the datetime metadataMark Wiebe2011-06-1711-275/+38
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The events don't fit well within the datetime, causing many special cases and complicating the datetime code. A better way to proceed, which received approval from Travis with regard to the reason events were introduced in the first place, is to do the following: * Generalize the structured arrays so they work with ufuncs * Either use NPY_UBYTE, or introduce a new integer modulo N type * The structured dtype [('date', 'M8[D]'), ('event', 'u8')], then will operate identically to the design of events in datetime. These other changes will have much use elsewhere as well.
| * | ENH: datetime-casting: Add casting parameters to more conversion functionsMark Wiebe2011-06-176-134/+493
| | | | | | | | | | | | | | | Also clean up the tests to work with the stricter date unit vs time unit boundary that has been added.
| * | ENH: datetime-string: Add support for datetime.tzinfo in datetime_as_stringMark Wiebe2011-06-166-185/+340
| | | | | | | | | | | | Also add some tests using the pytz library.
| * | ENH: datetime-autounit: Detect unit from string arrays as wellMark Wiebe2011-06-1610-1287/+1490
| | | | | | | | | | | | Also move datetime string functions to their own source file.
| * | ENH: datetime-parsing: Apply the datetime unit casting rules to string parsingMark Wiebe2011-06-164-54/+75
| | |
| * | ENH: datetime-casting: Refine the casting rules with a date vs time unit barrierMark Wiebe2011-06-164-80/+294
| | |
* | | TST: fix py3k failure for arrayprint. Closes #1875. Thanks to Derek Homeier.Ralf Gommers2011-06-211-1/+10
| |/ |/|
* | BUG[f2py]: fix --include_paths bug. Deprecated --include_paths in favor of ↵Pearu Peterson2011-06-212-10/+26
| | | | | | | | --include-paths. Updated docs.
* | Merge branch 'sortlib'Charles Harris2011-06-2010-530/+893
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * sortlib: STY: Delete commented out lines and add some comments. ENH: Generate the npysort library for linking during build, but do not install it. ENH: Make npy_sort.h private until the proper interface gets sorted out. ENH: Remove the _sort module. ENH: Directly initialize arraytypes with the sorting functions in the sort library. STY: Rename sort functions again, omit the npy_ bit from the suffix. ENH: Rename the sort functions using lower case suffix instead of upper case prefix. The name are now of the form <sortkind>_<npy_type>. ENH: Make scons build the npysort library. ENH: Move sorting functions into a library. STY: Move npy_intp definitions and associated formats into npy_common.h. STY: Reorganize core/setup.py a bit. ENH: Break out the sort functions into a separate file. ENH: Rename the sorting directory to npysort. ENH: Move comparison functions into an include file. Make include file for the future sorting library. Use prefixed numpy types. ENH: Rename _sortmodule.c.src and move it into a new sorting directory. Fix the build to deal with the new name and location.
| * | STY: Delete commented out lines and add some comments.Charles Harris2011-06-202-4/+4
| | |
| * | ENH: Generate the npysort library for linking during build, but do notCharles Harris2011-06-193-31/+5
| | | | | | | | | | | | install it.
| * | ENH: Make npy_sort.h private until the proper interface gets sorted out.Charles Harris2011-06-173-5/+3
| | |
| * | ENH: Remove the _sort module.Charles Harris2011-06-166-153/+3
| | |
| * | ENH: Directly initialize arraytypes with the sorting functions in the sortCharles Harris2011-06-163-2/+27
| | | | | | | | | | | | library.
| * | STY: Rename sort functions again, omit the npy_ bit from the suffix.Charles Harris2011-06-163-153/+151
| | |
| * | ENH: Rename the sort functions using lower case suffix instead of upperCharles Harris2011-06-164-186/+195
| | | | | | | | | | | | case prefix. The name are now of the form <sortkind>_<npy_type>.
| * | ENH: Make scons build the npysort library.Charles Harris2011-06-161-2/+18
| | |
| * | ENH: Move sorting functions into a library.Charles Harris2011-06-167-88/+230
| | |
| * | STY: Move npy_intp definitions and associated formats into npy_common.h.Charles Harris2011-06-165-110/+117
| | |
View Lorry Controller Status