[Numpy-svn] [numpy/numpy] 7c4622: ENH: added check for scalar array in get_ufunc_arg...

GitHub noreply@github....
Wed Jul 10 12:00:37 CDT 2013


  Branch: refs/heads/master
  Home:   https://github.com/numpy/numpy
  Commit: 7c46223294001328bae117ce4a04d3a1f53a1ad6
      https://github.com/numpy/numpy/commit/7c46223294001328bae117ce4a04d3a1f53a1ad6
  Author: Arink Verma <arinkverma@gmail.com>
  Date:   2013-07-06 (Sat, 06 Jul 2013)

  Changed paths:
    M numpy/core/src/umath/ufunc_object.c

  Log Message:
  -----------
  ENH: added check for scalar array in get_ufunc_arguments from ufunc_object.c
Check if object is of array dimension zero, then avoid heavy calls like PyArray_FromAny. Hence passing obj to out_op PyArrayObject unchanged. Also PyArray_FromAny(obj,NULL,0,0,0,NULL) calls PyArray_FromArray(obj,NULL,0) which return original obj


  Commit: 55bc296b5b415f4ca575e9d17ff930b14a89306d
      https://github.com/numpy/numpy/commit/55bc296b5b415f4ca575e9d17ff930b14a89306d
  Author: Arink Verma <arinkverma@gmail.com>
  Date:   2013-07-08 (Mon, 08 Jul 2013)

  Changed paths:
    M numpy/core/src/multiarray/ctors.c
    M numpy/core/src/umath/ufunc_object.c
    M numpy/core/src/umath/ufunc_type_resolution.c

  Log Message:
  -----------
  ENH: added check for scalar array in PyArray_FromArray from ctors.c
Check if object is of array dimension zero with newtype is NULL. If so return it directly instead of checking for casting.


  Commit: 9b3e719b8ab153335bcc9064cf186c823367bdbe
      https://github.com/numpy/numpy/commit/9b3e719b8ab153335bcc9064cf186c823367bdbe
  Author: Arink Verma <arinkverma@gmail.com>
  Date:   2013-07-08 (Mon, 08 Jul 2013)

  Changed paths:
    M numpy/core/src/umath/ufunc_type_resolution.c

  Log Message:
  -----------
  ENH: added check for scalar array in PyArray_FromArray from ctors.c
Check if object is of array dimension zero with newtype is NULL. If so return it directly instead of checking for casting.


  Commit: 021d52a1294c662c7e31918db2e26bdc10e37f20
      https://github.com/numpy/numpy/commit/021d52a1294c662c7e31918db2e26bdc10e37f20
  Author: Arink Verma <arinkverma@gmail.com>
  Date:   2013-07-08 (Mon, 08 Jul 2013)

  Changed paths:
    M numpy/core/src/multiarray/ctors.c

  Log Message:
  -----------
  ENH: added check for scalar array in PyArray_FromArray from ctors.c
Check if object is of array with newtype is NULL. If so return it directly instead of checking for casting.


  Commit: 4c3ad43fc6cf68813bea09e1451f371f0a992369
      https://github.com/numpy/numpy/commit/4c3ad43fc6cf68813bea09e1451f371f0a992369
  Author: Arink Verma <arinkverma@gmail.com>
  Date:   2013-07-10 (Wed, 10 Jul 2013)

  Changed paths:
    M numpy/core/src/multiarray/ctors.c

  Log Message:
  -----------
  ENH: added check for scalar array in PyArray_FromArray from ctors.c
Check if object is of array with newtype is NULL. If so return it directly instead of checking for casting.


  Commit: 6815f865678e88dab68c00c34fb1c9e7166f8f3c
      https://github.com/numpy/numpy/commit/6815f865678e88dab68c00c34fb1c9e7166f8f3c
  Author: Arink Verma <arinkverma@gmail.com>
  Date:   2013-07-10 (Wed, 10 Jul 2013)

  Changed paths:
    M numpy/core/src/multiarray/convert_datatype.c

  Log Message:
  -----------
  ENH: check for the same object in can_cast_scalar_to from convert_datatype.c
Check if two dtypes references to the same object in can_cast_scalar_to from convert_datatype.c,
then the result is always OK


  Commit: 12c5fc01e37d2235713c59ac0b1346294a8e8688
      https://github.com/numpy/numpy/commit/12c5fc01e37d2235713c59ac0b1346294a8e8688
  Author: Charles Harris <charlesr.harris@gmail.com>
  Date:   2013-07-10 (Wed, 10 Jul 2013)

  Changed paths:
    M numpy/core/src/multiarray/convert_datatype.c
    M numpy/core/src/multiarray/ctors.c
    M numpy/core/src/umath/ufunc_object.c

  Log Message:
  -----------
  Merge pull request #3509 from arinkverma/gsoc_performance

ENH: added short path scalar array in get_ufunc_arguments from ufunc_object.c


Compare: https://github.com/numpy/numpy/compare/49a587cd7862...12c5fc01e37d


More information about the Numpy-svn mailing list