Jason Grout wrote:
> Dag Sverre Seljebotn wrote:
>> Jason Grout wrote:
>>> ghtdak wrote:
>>>> On Jun 7, 11:29 am, Dag Sverre Seljebotn <da...@student.matnat.uio.no>
>>>> wrote:
>>>>> Glenn Tarbox, PhD wrote:
>>>>>> setup.py is pulled directly from the cython tutorial docs:
>>>>>> tar...@puget:$ cat setup.py
>>>>>> from distutils.core import setup
>>>>>> from distutils.extension import Extension
>>>>>> from Cython.Distutils import build_ext
>>>>>> setup(
>>>>>> cmdclass = {'build_ext': build_ext},
>>>>>> ext_modules = [Extension("testnumpy", ["testnumpy.pyx"])]
>>>>>> )
>>>>>> The numpy include directories were problematic previously (an extra
>>>>>> "core" directory level recently introduced) although Trac indicates
>>>>>> that issue was fixed and I'm "pretty sure" I first saw the include
>>>>>> problem in the notebook which doesn't occur.
>>>>> Note that that setup.py doesn't include NumPy headers, which are usually
>>>>> needed if numpy is cimported. Please do
>>>>>
>>>>> import numpy
>>>>> ...
>>>>> setup(
>>>>> ...
>>>>> ext_modules = [Extension(..., include_path=[numpy.get_include()])]
>>>>> )
>>>>>
>>>> cool. a little reading / poking around and came up with:
>>>>
>>>>
>>>> tar...@puget:$ cat setup.py
>>>> from distutils.core import setup
>>>> from distutils.extension import Extension
>>>> from Cython.Distutils import build_ext
>>>> import numpy
>>>>
>>>> setup(
>>>> cmdclass = {'build_ext': build_ext},
>>>> ext_modules = [Extension("testnumpy",
>>>> ["testnumpy.pyx"],include_dirs=[numpy.get_include()])]
>>>> )
>>> I thought #6053 fixed the notebook numpy issue. Is there more to the
>>> issue than what is done there? Are the tests above being done on 4.0
>>> (e.g., sagenb.org)?
>>>
>>>
>>> The following works fine on sage 4.0 (e.g., sagenb.org):
>>>
>>> %cython
>>> cimport numpy as np
>>> import numpy as np
>>> DTYPE = np.int
>>> ctypedef np.npy_intp DTYPE_t
>>>
>>> Note that I had to change the ctypedef. There is no int_t defined in
>>> the numpy.pxd file (devel/sage/sage/ext/numpy.pxd). Should there be an
>>> np.int_t type?
>> Hmm. Strange. Do you use the numpy.pxd shipped with Cython? You should
>> -- it evolves together with features in Cython, while the one shipping
>> with NumPy itself contains no support for things that has been going on
>> beyond Pyrex.
>>
>> int_t should be in there, see e.g.
>>
>> http://hg.cython.org/cython-devel/file/51fa7e425dc8/Cython/Includes/numpy.pxd#l260
>>
>> IIRC it's been there since the beginning.
>>
>
> No, we don't. We should, though. This is
> http://trac.sagemath.org/sage_trac/ticket/4571. Any volunteers?
If there is, please consider pushing the patch to Cython, and then it
will likely get into Sage the next time Sage upgrades Cython. Sage is
usually built and tested as part of the Cython release process anyway.
--
Dag Sverre
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
