summaryrefslogtreecommitdiff
path: root/scipy/base
Commit message (Collapse)AuthorAgeFilesLines
...
* Comment cleanup.chanley2005-11-021-3/+0
|
* CJH -- 11/2/2005 -- Modified the nbytes dictionary to return values in bytes ↵chanley2005-11-021-1/+5
| | | | instead of bits. Makes the assumption that 8 bits is 1 byte.
* added support for __array_struct__Travis Oliphant2005-11-023-1/+96
|
* Added nbytes attribute and nbytes dictionaryTravis Oliphant2005-11-013-5/+39
|
* Fixed Concatenate to use new ArrayType interface so flexible arrays are ↵Travis Oliphant2005-11-012-10/+9
| | | | supported.
* Reapplied patch to interpret sequences more cautiously.edschofield2005-11-011-2/+11
|
* Reverted the PySequence_Check patch. DOK matrices should be fixed instead.edschofield2005-11-011-10/+1
|
* Changed array_fromobject() to treat failure of the Array_FromSequence() call edschofield2005-11-011-4/+13
| | | | | | | | | | differently. Now it clears the error flag and tries interpreting the object as a scalar. This allows more graceful error handling of badly behaved objects like DOK matrices that pass PySequence_Check but can't be indexed like sequences. Fixed small typos.
* Added ISALIGNED macro and documentation update.Travis Oliphant2005-11-012-0/+2
|
* Fixed to allow reflected operators instead of OBJECT addition.Travis Oliphant2005-11-015-17/+58
|
* r707@803636f3: kern | 2005-10-31 13:19:55 -0800Robert Kern2005-10-311-1/+1
| | | | Fix put() typo
* Added more attribute access to flags object.Travis Oliphant2005-10-311-0/+50
|
* Allow flags dictionary attribute accessTravis Oliphant2005-10-311-0/+50
|
* Added bool8 type.Travis Oliphant2005-10-311-0/+2
|
* Fixed small typoedschofield2005-10-311-1/+1
|
* Don't use deconvolve for polydiv, as that requires scipy.signal to be installed.cookedm2005-10-301-1/+9
| | | | Do it using synthetic division. Could be more efficient.
* Don't try to handle ImportError in deconvolve when scipy.signal doesn't existcookedm2005-10-301-4/+1
|
* Fixed nx -> NXTravis Oliphant2005-10-301-2/+2
|
* Change to using PY_ARRAY_TYPES_PREFIX instead of PYARRAY_TYPES_PREFIX, tocookedm2005-10-302-6/+6
| | | | agree with PY_ARRAY_UNIQUE_SYMBOL.
* Make it easier for extension writers. If PYARRAY_TYPES_PREFIX is definedcookedm2005-10-302-20/+119
| | | | | | | | | to some value before arrayobject.h or ufuncobject.h is included, that value will be prefixed before types that typedef'd. So, if PYARRAY_TYPES_PREFIX is defined to be PA_, then, for instance, Bool as a type will not exist, but instead PA_Bool will have been defined. There is no problem with name mismatches as typedefs are not used in linking.
* some clean up for scimath.py, use type objects instead of typecharscookedm2005-10-291-14/+13
|
* small whitespacecookedm2005-10-291-1/+1
|
* Convert polynomial to use type objects instead of old Numeric typecharscookedm2005-10-291-22/+20
|
* TLC for polynomial.py: cleanup, poly1d class is now new-type, and unittestscookedm2005-10-292-3/+74
| | | | The unit tests are doctests.
* Give ufunclike some TLC. Cleanup, add docstrings and unit tests.cookedm2005-10-292-13/+91
| | | | | The unit tests are as doctests, and currently use the deprecated test_suite method of running.
* Whitespace cleanup and make classes new-typecookedm2005-10-299-121/+112
|
* Altered vectorize to allow default arguments in function.Travis Oliphant2005-10-291-4/+11
|
* More cosmetic changes.Travis Oliphant2005-10-291-10/+10
|
* Added PyUFunc_On_Om to UFunc C-APITravis Oliphant2005-10-295-5/+9
|
* Some refactoringcookedm2005-10-281-111/+119
| | | | | | - poly1d is a new-style class now - no import * now - whitespace, prettying
* Cosmetic changes mainly. Remove unused itemsize from scanfunc arguments.Travis Oliphant2005-10-284-10/+8
|
* change PyArray_Scalar to take a void * instead of char *, and some ↵cookedm2005-10-283-54/+33
| | | | whitespace cleanup
* Change PyArray_Scalar to take a void * instead of char *cookedm2005-10-281-2/+2
|
* Whitespace cleanupcookedm2005-10-281-133/+133
|
* Add __all__ statements to base.{oldnumeric,numeric,numerictypes}cookedm2005-10-283-255/+260
| | | | This should cut down the number of accidental imports
* Fixed SimpleNewFromData to use correct flags.Travis Oliphant2005-10-281-1/+1
|
* Fixed complex to bool issues.Travis Oliphant2005-10-283-24/+108
|
* Introduced PyArray_SimpleNewFromData as a replacement to PyArray_FromDimsAndDataPearu Peterson2005-10-281-0/+2
|
* Removed unused requirement that compare deal with misbehaved arrays.Travis Oliphant2005-10-282-41/+16
|
* Small cleanups in arrayobject.hTravis Oliphant2005-10-281-6/+6
|
* Altered default flags in PyArray_BufferConverter and fixed signbit and friends.Travis Oliphant2005-10-282-2/+2
|
* Fixed problem with correlate.Travis Oliphant2005-10-275-27/+47
|
* Fix 0-d array indexing.Travis Oliphant2005-10-273-4/+19
|
* Fix for rank-0 arrays. Now a[0] = 1 throws an IndexError, just likeedschofield2005-10-271-1/+6
| | | | | | | reading from a[0] does. ** Note: assigning a[-1] = 0 still segfaults! **
* More docstring fixes for mean() and sum().edschofield2005-10-272-43/+35
|
* Trivial changes to a.mean() docstringedschofield2005-10-271-2/+2
|
* Improved docstrings:edschofield2005-10-273-190/+261
| | | | | | | | | * Fuller explanation of default casting behaviour of a.sum(), a.mean(), scipy.sum(), and scipy.average() * Changed some array variable names from 'x' to 'a' in function_base.py for consistency with docstrings in arraymethods.c * Other cosmetic fixes to docstrings and code formatting
* Made ScalarType tuple so that it could be 2nd argument to isinstance.Pearu Peterson2005-10-271-0/+1
|
* Fixed stepping issue for length-1 arrays.Travis Oliphant2005-10-271-1/+1
|
* Fixed problem with temporary allocation in PyArray_Transpose.Travis Oliphant2005-10-261-5/+5
|
View Lorry Controller Status