Hi,
I have installed numpy but the unit tests fail. When I ran them, I got
Traceback (most recent call last):
File "/home/jhtu/local/lib/python2.7/site-packages/numpy/testing/
decorators.py", line 215, in knownfailer
return f(*args, **kwargs)
File "/home/jhtu/local/lib/python2.7/site-packages/numpy/core/tests/
test_umath_complex.py", line 312, in test_special_values
assert_almost_equal(np.log(np.conj(xa[i])), np.conj(np.log(xa[i])))
File "/home/jhtu/local/lib/python2.7/site-packages/numpy/testing/
utils.py", line 443, in assert_almost_equal
raise AssertionError(msg)
AssertionError:
Arrays are not almost equal
ACTUAL: array([-inf+3.14159265j])
DESIRED: array([-inf-3.14159265j])
This was with numpy built against MKL. To install I modified site.cfg
to read
[mkl]
library_dirs = /opt/intel/mkl/10.2.4.032/lib/em64t
include_dirs = /opt/intel/mkl/10.2.4.032/include
lapack_libs = mkl_lapack
mkl_libs = mkl_intel_lp64, mkl_intel_thread, mkl_core
My cluster is using Intel Xeon processors, and I edited cc_exe as
follows
cc_exe = 'icc -O2 -fPIC'
I installed using
python setup.py config --compiler=intel build_clib --compiler=intel
build_ext --compiler=intel install --prefix=/home/jhtu/local
Jonathan Tu
On Jan 18, 2011, at 3:39 PM, Ilan Schnell wrote:
> The MKL configuration looks right, except that I had to use:
> mkl_libs = mkl_intel_lp64, mkl_intel_thread, mkl_core, iomp5
>
> During the build process, it should tell you what it is linking
> aginast. Look at the compiler options passed to icc.
>
> - Ilan
>
> On Tue, Jan 18, 2011 at 2:31 PM, Jonathan Tu <[email protected]>
> wrote:
>> Hi,
>>
>> I realized that my cluster has MKL installed. I've been trying to
>> install against MKL, but am having trouble getting this to work.
>> After it finishes, I do
>>
>> import numpy
>> numpy.show_config()
>>
>> and nothing about the MKL libraries shows up. I have edited site.cfg
>> to read like this:
>>
>> [mkl]
>> library_dirs = /opt/intel/mkl/10.2.4.032/lib/em64t
>> include_dirs = /opt/intel/mkl/10.2.4.032/include
>> lapack_libs = mkl_lapack
>> mkl_libs = mkl, guide
>>
>> My cluster is using Intel Xeon processors, and I edited cc_exe as
>> follows
>>
>> cc_exe = 'icc -O2 -fPIC'
>>
>> Then I did
>>
>> python setup.py config --compiler=intel build_clib --compiler=intel
>> build_ext --compiler=intel install --prefix=/home/jhtu/local
>>
>>
>>
>>
>> Jonathan Tu
>>
>>
>>
>>
>> On Jan 18, 2011, at 3:28 PM, Ilan Schnell wrote:
>>
>>> Hello Jonathan,
>>>
>>> yes, numpy work fine under Python 2.7 now. I don't see why building
>>> numpy against the system ATLAS should not work, as long as you
>>> install the developer version with the header files, and make sure
>>> that
>>> you edit the site.cfg file correct.
>>>
>>> - Ilan
>>>
>>> On Tue, Jan 18, 2011 at 10:39 AM, Jonathan Tu <[email protected]>
>>> wrote:
>>>> Hi,
>>>>
>>>> I need to reinstall numpy because the cluster I am using was
>>>> recently
>>>> overhauled. I am wondering if numpy works with Python 2.7 now.
>>>>
>>>> Also, I would like numpy to run as fast as possible. The last
>>>> time I
>>>> did this, I was advised to install ATLAS by hand, as the one that
>>>> comes with RHEL is not suitable. The first time I tried this, I
>>>> kept
>>>> running into problems that I think were due to mismatched fortran
>>>> compilers. Is there a good resource for how to do this? I am
>>>> fairly
>>>> new to Linux.
>>>>
>>>>
>>>>
>>>> Thanks,
>>>>
>>>>
>>>>
>>>> Jonathan Tu
>>>> _______________________________________________
>>>> NumPy-Discussion mailing list
>>>> [email protected]
>>>> http://mail.scipy.org/mailman/listinfo/numpy-discussion
>>>>
>>> _______________________________________________
>>> NumPy-Discussion mailing list
>>> [email protected]
>>> http://mail.scipy.org/mailman/listinfo/numpy-discussion
>>
>> _______________________________________________
>> NumPy-Discussion mailing list
>> [email protected]
>> http://mail.scipy.org/mailman/listinfo/numpy-discussion
>>
> _______________________________________________
> NumPy-Discussion mailing list
> [email protected]
> http://mail.scipy.org/mailman/listinfo/numpy-discussion
_______________________________________________
NumPy-Discussion mailing list
[email protected]
http://mail.scipy.org/mailman/listinfo/numpy-discussion
