Permalink
Commits on Aug 11, 2013
  1. Merge pull request #3599 from charris/fix-operand-flags-test

    BUG: Fix test_operand_flags test.
    charris committed Aug 11, 2013
  2. Merge pull request #3594 from juliantaylor/is_aligned_maint

    MAINT: move npy_is_aligned to common.h
    charris committed Aug 11, 2013
  3. Merge pull request #3526 from zyv/fix_int_inheritance

    BUG: Fix PyType_FastSubclass() check for numpy.int_ and Py3k inheritance
    charris committed Aug 11, 2013
  4. BUG: Fix test_operand_flags test.

    The test tests an inner loop in operand_flag_tests.c.src that expects
    a long type, but it is tested using 'i8'. This fails when long is not
    'i8'.
    
    Closes #3363.
    charris committed Aug 11, 2013
Commits on Aug 10, 2013
  1. TST: Add regression test to check numpy.int_ inheritance on Py2/Py3

    Signed-off-by: Yury V. Zaytsev <yury@shurup.com>
    zyv committed Jul 17, 2013
  2. BUG: Document that numpy.int_ doesn't inherit from int on Py3k

    Signed-off-by: Yury V. Zaytsev <yury@shurup.com>
    zyv committed Jul 16, 2013
  3. BUG: Fix PyType_FastSubclass() check for numpy.int_

    Signed-off-by: Yury V. Zaytsev <yury@shurup.com>
    zyv committed Jul 16, 2013
  4. Merge pull request #3517 from juliantaylor/memchr-all-any

    ENH: use memchr for unit stride all/any
    charris committed Aug 10, 2013
  5. Merge pull request #3596 from ecprice/a

    Fix documentation description of loop spam in interactive interpreter
    charris committed Aug 10, 2013
  6. Merge pull request #3597 from rgommers/testing-module-cleanup

    MAINT: remove unused and broken parts of numpy.testing
    charris committed Aug 10, 2013
  7. MAINT: remove unused and broken parts of numpy.testing

    Deprecate np.testing.importall - it's pointless and partially broken.
    rgommers committed Aug 10, 2013
Commits on Aug 9, 2013
  1. ENH: add NPY_LIKELY and NPY_UNLIKELY macros for branching hints

    Use it for npy_is_aligned expecting alignments of the power of two.
    Cuts down the time spent in _IsAligned by the testsuite relative
    to the rest of multiarray.so from 0.6% to 0.4%
    juliantaylor committed Aug 9, 2013
  2. MAINT: move npy_is_aligned to common.h

    also make it and some other alignment operators a bit faster by using
    unsigned integers and bitwise and if possible.
    Make use of npy_is_aligned in _IsAligned too.
    juliantaylor committed Aug 9, 2013
  3. ENH: improve numpy.all()/any()

    Unroll the loop once and use pminub/pmaxub to save a slow pmovmskb
    instruction. Improves performance by 50% on some AMD chips.
    Also add a pure libc path using memcmp and memchr for non amd64 systems.
    The libc path can be faster with a very modern cpu and libc version,
    e.g. an i7 with glibc 2.17 is about 20% faster than our code but many
    other tested platforms are much slower (2.12 xeon, core2duo) or same
    speed (2.17 phenom).
    The numpy code can be removed in future when faster libc versions and
    cpus are more commonly available.
    juliantaylor committed Jul 11, 2013
Commits on Aug 5, 2013
  1. Merge pull request #3513 from seberg/0d-linalg

    ENH: inv/solve work with empty inner and others empty outer array
    charris committed Aug 5, 2013
  2. Merge pull request #3565 from charris/documentation-fixes

    Documentation fixes for `basics.io.genfromtxt.rst` and `creation.py`
    charris committed Aug 5, 2013
  3. Merge pull request #3521 from arinkverma/gsoc_performance

    ENH: Avoiding NPY_BEGIN_THREADS for small arrays can speed-up trivial_three_operand_loop by 5%
    charris committed Aug 5, 2013
  4. Merge pull request #3575 from seberg/issue-3458

    BUG: Boolean assignment allowed writing to 0-sized array
    charris committed Aug 5, 2013
  5. Merge pull request #3576 from mspacek/patch-3

    MAINT: fix typo in npy-format.txt
    charris committed Aug 5, 2013
  6. BUG: Boolean assignment allowed writing to 0-sized array

    This was because of the assumption that broadcasting works
    if the dimension is not > 1, but correct is != 1.
    
    Adepted from a patch provided by prossahl.
    
    Closes gh-3458.
    seberg committed Aug 5, 2013
  7. Merge pull request #3568 from mspacek/patch-1

    Fix doc typo
    charris committed Aug 5, 2013
  8. ENH: For smaller array, added macro NPY_BEGIN_THREADS_THRESHOLDED in …

    …ndarraytypes.h
    
    Avoiding NPY_BEGIN_THREADS for small arrays, can speed-up trivial_three_operand_loop by 5%.
    As releases of GIL, then quickly restoring just after small operation doesn't benefit.
    arinkverma committed Jul 12, 2013
  9. Merge pull request #3572 from pv/fix-runtests

    MAINT: runtests: make it work on Fedora
    njsmith committed Aug 5, 2013
Commits on Aug 4, 2013
  1. ENH: inv/solve work with empty inner and others empty outer array

    This makes the inverse of a 0x0 array simply be 0x0 again. It
    also modifies the no-empty array check in favor of a no-empty
    *inner* array, since the gufuncs seem to handle the other case
    fine.
    seberg committed Jul 10, 2013
  2. Fix doc typo

    mspacek committed Aug 4, 2013
Commits on Aug 2, 2013
  1. Merge pull request 'gh-3497'

    * gh-3497:
      BUG: Make np.insert check for out of bounds axis arguments.
      TST: add test for negative axis values in np.insert.
      BUG: Fix bug in np.insert when axis=-1
    charris committed Aug 2, 2013
  2. BUG: Make np.insert check for out of bounds axis arguments.

    Also add test for IndexError exception when axis is out of bounds.
    FelixHartmann committed with charris Jul 9, 2013