numpy 1.04 numpy.test() hang
Good morning. I just installed the Windows binary of numpy 1.04. When I ran numpy.test() in IDLE (the Python shell that comes with Python), the program hang (or at least is running for half an hour). I am using Windows XP, duel core intel CPU. Does anyone know what is going on? Thanks, Geoffrey
Geoffrey Zhu wrote:
Good morning.
I just installed the Windows binary of numpy 1.04. When I ran numpy.test() in IDLE (the Python shell that comes with Python), the program hang (or at least is running for half an hour). I am using Windows XP, duel core intel CPU.
Does anyone know what is going on?
No. Run numpy.test(verbosity=2) so it will print out each test name before running the test. Then we might have some idea about where the hang is coming from. -- Robert Kern "I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco
On Thu, 08 Nov 2007 12:12:42 -0600 Robert Kern <robert.kern@gmail.com> wrote:
Geoffrey Zhu wrote:
Good morning.
I just installed the Windows binary of numpy 1.04. When I ran numpy.test() in IDLE (the Python shell that comes with Python), the program hang (or at least is running for half an hour). I am using Windows XP, duel core intel CPU.
Does anyone know what is going on?
No. Run numpy.test(verbosity=2) so it will print out each test name before running the test. Then we might have some idea about where the hang is coming from.
-- Robert Kern
"I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
I can confirm the problem on an AMD x86_64 with python2.5.1. The test hangs at check_cdouble (numpy.tests.test_linalg.test_det) Nils
On Nov 8, 2007 12:12 PM, Robert Kern <robert.kern@gmail.com> wrote:
Geoffrey Zhu wrote:
Good morning.
I just installed the Windows binary of numpy 1.04. When I ran numpy.test() in IDLE (the Python shell that comes with Python), the program hang (or at least is running for half an hour). I am using Windows XP, duel core intel CPU.
Does anyone know what is going on?
No. Run numpy.test(verbosity=2) so it will print out each test name before running the test. Then we might have some idea about where the hang is coming from.
-- Robert Kern
"I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Thanks for the hint, Robert. It hangs on check_cdouble (numpy.tests.test_linalg.test_det). Also testip_not_allclose() had three warnings. I guess that's probably okay. testip_not_allclose (numpy.core.tests.test_numeric.test_allclose_inf) ... Warning: invalid value encountered in absolute Warning: invalid value encountered in absolute Warning: invalid value encountered in less_equal ok Thanks, Geoffrey
Geoffrey Zhu wrote:
On Nov 8, 2007 12:12 PM, Robert Kern <robert.kern@gmail.com> wrote:
Geoffrey Zhu wrote:
Good morning.
I just installed the Windows binary of numpy 1.04. When I ran numpy.test() in IDLE (the Python shell that comes with Python), the program hang (or at least is running for half an hour). I am using Windows XP, duel core intel CPU.
Does anyone know what is going on?
No. Run numpy.test(verbosity=2) so it will print out each test name before running the test. Then we might have some idea about where the hang is coming from.
-- Robert Kern
"I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Thanks for the hint, Robert.
It hangs on check_cdouble (numpy.tests.test_linalg.test_det).
Also testip_not_allclose() had three warnings. I guess that's probably okay.
testip_not_allclose (numpy.core.tests.test_numeric.test_allclose_inf) ... Warning: invalid value encountered in absolute Warning: invalid value encountered in absolute Warning: invalid value encountered in less_equal ok
Are you on x86-64, too ? Which BLAS are you using ? This smells like a C/Fortran problem (because it happens with complex values only). cheers, David
On Nov 8, 2007 10:06 PM, David Cournapeau <david@ar.media.kyoto-u.ac.jp> wrote:
Geoffrey Zhu wrote:
On Nov 8, 2007 12:12 PM, Robert Kern <robert.kern@gmail.com> wrote:
Geoffrey Zhu wrote:
Good morning.
I just installed the Windows binary of numpy 1.04. When I ran numpy.test() in IDLE (the Python shell that comes with Python), the program hang (or at least is running for half an hour). I am using Windows XP, duel core intel CPU.
Does anyone know what is going on?
No. Run numpy.test(verbosity=2) so it will print out each test name before running the test. Then we might have some idea about where the hang is coming from.
-- Robert Kern
"I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Thanks for the hint, Robert.
It hangs on check_cdouble (numpy.tests.test_linalg.test_det).
Also testip_not_allclose() had three warnings. I guess that's probably okay.
testip_not_allclose (numpy.core.tests.test_numeric.test_allclose_inf) ... Warning: invalid value encountered in absolute Warning: invalid value encountered in absolute Warning: invalid value encountered in less_equal ok
Are you on x86-64, too ? Which BLAS are you using ? This smells like a C/Fortran problem (because it happens with complex values only).
cheers,
David
_______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Hi David, Although the processor is Intel Duo Core E6700, which supports x86-64, I am only using 32-bit Windows XP. I am not using my own BLAS. I am simply using the pre-compiled binary. I installed 1.04 over the old 1.031. I don't know if it is caused by some files that are left behind. Thanks, Geoffrey
On Nov 9, 2007 10:14 AM, Geoffrey Zhu <zyzhu2000@gmail.com> wrote:
On Nov 8, 2007 10:06 PM, David Cournapeau <david@ar.media.kyoto-u.ac.jp> wrote:
Geoffrey Zhu wrote:
On Nov 8, 2007 12:12 PM, Robert Kern <robert.kern@gmail.com> wrote:
Geoffrey Zhu wrote:
Good morning.
I just installed the Windows binary of numpy 1.04. When I ran numpy.test() in IDLE (the Python shell that comes with Python), the program hang (or at least is running for half an hour). I am using Windows XP, duel core intel CPU.
Does anyone know what is going on?
No. Run numpy.test(verbosity=2) so it will print out each test name before running the test. Then we might have some idea about where the hang is coming from.
-- Robert Kern
"I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Thanks for the hint, Robert.
It hangs on check_cdouble (numpy.tests.test_linalg.test_det).
Also testip_not_allclose() had three warnings. I guess that's probably okay.
testip_not_allclose (numpy.core.tests.test_numeric.test_allclose_inf) ... Warning: invalid value encountered in absolute Warning: invalid value encountered in absolute Warning: invalid value encountered in less_equal ok
Are you on x86-64, too ? Which BLAS are you using ? This smells like a C/Fortran problem (because it happens with complex values only).
cheers,
David
_______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Hi David,
Although the processor is Intel Duo Core E6700, which supports x86-64, I am only using 32-bit Windows XP. I am not using my own BLAS. I am simply using the pre-compiled binary. I installed 1.04 over the old 1.031. I don't know if it is caused by some files that are left behind.
Thanks, Geoffrey
Okay. I verified (by looking at which DLL python.exe actually loads) that it is using _dotblas.pyd under C:\Python25\Lib\site-packages\numpy\core that came with the pre-built 1.04 binary. It is not using any left-over from 1.031.
Very interesting! If I use the MSI file, numpy.test() hangs. If, however, I use the EGG file, it is actually fine.
On Fri, 9 Nov 2007 11:31:54 -0600 "Geoffrey Zhu" <zyzhu2000@gmail.com> wrote:
Very interesting! If I use the MSI file, numpy.test() hangs. If, however, I use the EGG file, it is actually fine. _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
How do I install an EGG ? I am testing Windows XP SP2. Nils
On Nov 9, 2007 11:42 AM, Nils Wagner <nwagner@iam.uni-stuttgart.de> wrote:
On Fri, 9 Nov 2007 11:31:54 -0600 "Geoffrey Zhu" <zyzhu2000@gmail.com> wrote:
Very interesting! If I use the MSI file, numpy.test() hangs. If, however, I use the EGG file, it is actually fine. _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
How do I install an EGG ? I am testing Windows XP SP2.
Nils
_______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
I downloaded and ran ez_setup.py from here: http://peak.telecommunity.com/DevCenter/EasyInstall . And then I downloaded the egg file and ran easy_install c:\mypath\numpy-1.0.4-py2.5-win32.egg. Do not put the egg file in any of your python package directories, or easy_install won't unpack it. I haven't figured out why.
On 11/9/07, Geoffrey Zhu <zyzhu2000@gmail.com> wrote:
And then I downloaded the egg file and ran easy_install c:\mypath\numpy-1.0.4-py2.5-win32.egg. Do not put the egg file in any of your python package directories, or easy_install won't unpack it. I haven't figured out why.
Most likely because "installing an egg" for the most part simply means putting it on your python path. I say mostly, because of the issue of unzipping that your experiencing as well as possibly adding it to your easy_install.pth file and maybe even an issue with the part of the python path the egg being on needing to be one that allows .pth files. This is the long answer (and should likely be considered a bug), the short answer is "don't do that". :) You can also install setuptools with the windows installer at: <http://pypi.python.org/pypi/setuptools>.
Geoffrey Zhu wrote:
Very interesting! If I use the MSI file, numpy.test() hangs. If, however, I use the EGG file, it is actually fine.
Can you find the md5sum of these files? There is a md5sum.exe at http://www.etree.org/md5com.html It would be good to verify that you have the correct files. Thanks, -Travis
On Nov 9, 2007 11:45 AM, Travis E. Oliphant <oliphant@enthought.com> wrote:
Geoffrey Zhu wrote:
Very interesting! If I use the MSI file, numpy.test() hangs. If, however, I use the EGG file, it is actually fine.
Can you find the md5sum of these files?
There is a md5sum.exe at
http://www.etree.org/md5com.html
It would be good to verify that you have the correct files.
Thanks,
-Travis
_______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Sure. Here they are: a3be796f61ee63c06675524f3dda8836 *numpy-1.0.4.win32-py2.5.msi 07bb650726caa946abfa464d8968437c *numpy-1.0.4-py2.5-win32.egg I downloaded them from sourceforge. By the way, something is gong on with the sourceforge download page. When I tried to download the MSI file, sourceforge always comes up with a connection error. I figured it was in the ajax and managed to click the direct download link really quickly. Another sign is that all the download counts for 1.04 are zero.
participants (6)
-
Alexander Michael -
David Cournapeau -
Geoffrey Zhu -
Nils Wagner -
Robert Kern -
Travis E. Oliphant