[Numpy-svn] [numpy/numpy] 367b40: BUG: core: ensure binop execution uses ufuncs as f...

[GitHub]

  Branch: refs/heads/master
  Home:   https://github.com/numpy/numpy
  Commit: 367b4094125062fa9018db0f310b28780259f29b
      https://github.com/numpy/numpy/commit/367b4094125062fa9018db0f310b28780259f29b
  Author: Pauli Virtanen <pav at iki.fi>
  Date:   2015-04-04 (Sat, 04 Apr 2015)

  Changed paths:
    M numpy/core/src/multiarray/arrayobject.c
    M numpy/core/src/multiarray/number.c

  Log Message:
  -----------
  BUG: core: ensure binop execution uses ufuncs as fallback

These changes only affect objects defining __numpy_ufunc__. Other
objects use the __array_priority__ mechanism to decide whether
NotImplemented should be returned or not.

The binops previously returned NotImplemented even if other._r<op>__ is
ndarray.__r<op>__, rather than trying to evaluate the result via ufuncs.
This causes problems in binops of two objects that both subclass from
ndarray and define __numpy_ufunc__.

The solution added here makes the total logic as follows:

    def __binop__(self, other):
  if (hasattr(other, '__numpy_ufunc__') and
      not isinstance(other, self.__class__) and
      hasattr(other, '__rop__') and
      other.__class__.__rop__ is not self.__class__.__rop__):
      return NotImplemented
  return np.binop(self, other)


  Commit: 6ba4fed3aa26979734faa462e172959fad69723b
      https://github.com/numpy/numpy/commit/6ba4fed3aa26979734faa462e172959fad69723b
  Author: Pauli Virtanen <pav at iki.fi>
  Date:   2015-04-04 (Sat, 04 Apr 2015)

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

  Log Message:
  -----------
  BUG: core: inplace ops don't have corresponding rhs ops, so no need to check the slot


  Commit: d1153b4cba5510bd1961e3a7ae7f0c07c124a202
      https://github.com/numpy/numpy/commit/d1153b4cba5510bd1961e3a7ae7f0c07c124a202
  Author: Marten van Kerkwijk <mhvk at astro.utoronto.ca>
  Date:   2015-04-04 (Sat, 04 Apr 2015)

  Changed paths:
    M numpy/core/tests/test_multiarray.py

  Log Message:
  -----------
  Add tests that __numpy_ufunc__ works with subclasses.

In particular, for a class S3(S2), where S2 defines __numpy_ufunc__, ensure
that, e.g., S3() + S2() does *not* pass control to S2 if S2 and S3 share the
same __radd__.


  Commit: eecb2e3c07f29c0ac991d364a846a2f8293a432a
      https://github.com/numpy/numpy/commit/eecb2e3c07f29c0ac991d364a846a2f8293a432a
  Author: Charles Harris <charlesr.harris at gmail.com>
  Date:   2015-05-05 (Tue, 05 May 2015)

  Changed paths:
    M numpy/core/src/multiarray/arrayobject.c
    M numpy/core/src/multiarray/number.c
    M numpy/core/tests/test_multiarray.py

  Log Message:
  -----------
  Merge pull request #5748 from mhvk/numpyufunc-binop-fix

BUG Numpy ufunc binop fix for subclasses with __numpy_ufunc__ (closes #4815)


Compare: https://github.com/numpy/numpy/compare/a29c296f0bf3...eecb2e3c07f2