summaryrefslogtreecommitdiff
path: root/numpy
Commit message (Collapse)AuthorAgeFilesLines
...
* 3K: core: fix a few str vs bytes issues in testsPauli Virtanen2010-02-213-14/+20
|
* 3K: matrixlib: adapt to changes in str.translatePauli Virtanen2010-02-211-14/+26
|
* 3K: ENH: regenerate mtrand.c with a newer Cython, for better Python 3 ↵Pauli Virtanen2010-02-211-4686/+5426
| | | | compatibility
* 3K: core/numerictypes: fix handling of Python types, and recognize bytes as ↵Pauli Virtanen2010-02-211-7/+15
| | | | a Python type
* 3K: BUG: fix unravel_index integer divisionPauli Virtanen2010-02-211-1/+1
|
* 3K: ENH: ensure integer scalar types are hashable -- we don't automatically ↵Pauli Virtanen2010-02-211-8/+12
| | | | inherit the hash implementation from Python ints on 3K
* DEP: Fix more files for unittest deprecated functions. It isn't clear why theseCharles Harris2010-02-2112-206/+206
| | | | tests didn't issue deprecation warnings. Are they being run?
* DEP: Fix deprecation warnings in Python 3.1. The warnings come from the unittestCharles Harris2010-02-2011-434/+434
| | | | | | | | | module. The fix should be good for Python >= 2.4 and used the following sed script: s/\<failUnless\>/assertTrue/g s/\<failIf\>/assertFalse/g s/\<failUnlessEqual\>/assertEqual/g s/\<failUnlessRaises\>/assertRaises/g
* BUG: ensure Py_TYPE is available in _dotblas.c on Python 2.4 and 2.5Pauli Virtanen2010-02-201-0/+6
|
* BUG: fix a few missing includes that may cause problems on 64-bitPauli Virtanen2010-02-203-0/+5
|
* 3K: BUG: fix refcount error in npy_PyFile_DupPauli Virtanen2010-02-201-0/+1
|
* BUG: properly check the return value of numpy.core._internal._dtype_from_pep3118Pauli Virtanen2010-02-201-3/+10
|
* 3K: BUG: core: fix ucsnarrow.c:MyPyUnicode_New to work correctly on Py3Pauli Virtanen2010-02-201-0/+3
|
* BUG: fix a crash due to mixing PyUnicode and PyStringPauli Virtanen2010-02-201-1/+10
|
* 3K: BUG: work around bugs in Python 3.1.1 2to3 by not using fixes_reducePauli Virtanen2010-02-2010-9/+42
| | | | Instead, manually import reduce where necessary.
* BUG: another single-file compilation fixPauli Virtanen2010-02-201-0/+2
|
* BUG: lib: ensure 'bytes' is imported in io.pyPauli Virtanen2010-02-201-1/+1
|
* 3K: core: fix single-file compilationPauli Virtanen2010-02-201-0/+1
|
* 3K: Some remaining Py_TYPE issuesPauli Virtanen2010-02-202-2/+2
|
* 3K: lib: more str vs bytes issues in the lib/io loadtxt, savetxt and genfromtxtPauli Virtanen2010-02-205-67/+104
|
* ENH: compat: additional Py3 convenience functionsPauli Virtanen2010-02-201-3/+9
|
* 3K: lib: fix savetxtPauli Virtanen2010-02-201-9/+23
| | | | | | This will make savetxt open files in the 'wb' mode on Python 3. To allow using any sort of newlines (which are different e.g. on Windows), add a new 'newline' keyword.
* 3K: lib: fix some bytes vs. str issues in _iotools.py and io.py -- mainly ↵Pauli Virtanen2010-02-205-88/+143
| | | | genfromtxt
* 3K: lib: even more bytes/str fixes in format.py testsPauli Virtanen2010-02-201-6/+8
|
* 3K: lib: use BytesIO in test_ioPauli Virtanen2010-02-201-109/+115
|
* 3K: lib: fix some bytes/str issues in _format.py and its testsPauli Virtanen2010-02-202-7/+22
|
* ENH: lib: write fortran-contiguous data to files using arr.T.tofile instead ↵Pauli Virtanen2010-02-201-4/+5
| | | | | | | | | of arr.data (required for Py3 compatibility) The issue is that when passing a buffer object to Python's io.BufferedWriter.write, it will try to obtain the buffer using PyBUF_ND | PyBUF_C_CONTIGUOUS. This will fail for strided arrays -- seems to be an issue in Python, as it probably should try to obtain a SIMPLE buffer.
* 3K: core: in tofile etc., ensure the file is flushed before switching to ↵Pauli Virtanen2010-02-201-0/+5
| | | | FILE* I/O
* 3K: lib: adapt safe_eval for Py3 ast modulePauli Virtanen2010-02-201-32/+88
|
* 3K: compat: make isfileobj recognize only real file objects with FILE* ↵Pauli Virtanen2010-02-201-1/+1
| | | | pointers as file objects
* 3K: lib: do not slice range() in _iotools needlesslyPauli Virtanen2010-02-201-1/+1
|
* 3K: lib: Make _datasource and its tests Py3 compatible + slight cleanup of ↵Pauli Virtanen2010-02-202-9/+12
| | | | the code
* 3K: fft: fix integer division in (i)fftshiftPauli Virtanen2010-02-201-2/+2
|
* 3K: f2py: use integer division to avoid problems with string multiplicationPauli Virtanen2010-02-201-1/+1
|
* 3K: f2py: break cyclic imports (which are not allowed on Py3) by moving them ↵Pauli Virtanen2010-02-202-2/+3
| | | | to functions
* 3K: ENH: core: alias __round__ -> round for the generic scalar type, for the ↵Pauli Virtanen2010-02-201-0/+6
| | | | round() builtin
* ENH: core: Fix the test for #99 -- np.long == raw Python long. The test ↵Pauli Virtanen2010-02-201-5/+5
| | | | should test np.intp instead
* 3K: core: fix pickling of unicode scalars on Py3 -- they do not have buffer ↵Pauli Virtanen2010-02-201-0/+8
| | | | interface, so we need to get the buffer via the standard route
* 3K: lib: bytes vs. str fixes in lib.format and lib.ioPauli Virtanen2010-02-202-5/+7
|
* 3K: core: use BytesIO instead of StringIO in test_regressions on Py3Pauli Virtanen2010-02-201-0/+4
|
* ENH: core: allow unicode file names in PyArray_DumpPauli Virtanen2010-02-201-1/+1
|
* 3K: core: allow unicode type strings in __array_interface__ since we allow ↵Pauli Virtanen2010-02-201-1/+12
| | | | those also in dtypes
* 3K: linalg: fix some str/bytes issuesPauli Virtanen2010-02-201-20/+30
|
* 3K: core: fix up some testsPauli Virtanen2010-02-201-2/+2
|
* 3K: ENH: core/numerictypes: Update string/unicode aliases for Py3Pauli Virtanen2010-02-201-15/+55
|
* 3K: ENH: on py3k, rename scalar type display names as unicode_ -> str_ and ↵Pauli Virtanen2010-02-201-1/+16
| | | | string_ -> bytes_
* BUG: core: make sure PyArray_FromAny respects wanted type and flags also for ↵Pauli Virtanen2010-02-202-2/+12
| | | | PEP 3118 objects
* ENH: implement PEP 3118 buffer consumer on Py2.6, adding a simple Memoryview ↵Pauli Virtanen2010-02-2010-16/+371
| | | | | | | | | | | | 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.
* 3K: ENH: make defchararray work on Py3, and make its tests Py3 compliantPauli Virtanen2010-02-202-80/+124
| | | | There were mainly issues in mixing bytes/unicode on Py3.
* ENH: Add some tools to numpy.compatPauli Virtanen2010-02-202-3/+16
|