Re: SciPy-user digest, Vol 1 #201 - 7 msgs
import scipy Traceback (most recent call last): File "<stdin>", line 1, in ? File "/home/fgibbons/lib/python2.1/site-packages/scipy/__init__.py",
Travis, I tried this, and it does allow me to finish building, but why are there no files in the xplt directory (except for setup_xplt.py), when I can see them on the CVS page. I see that someone else had a similar problem to me, with the downloaded tarball. Thanks for the tip. However.... I realized that I needed to get the full LAPACK installation (and therefore had to also build BLAS since I have no vendor-supplied version). My full LAPACK weighs in at a healthy 6MB, and I followed the procedures in the installation guide to replace ATLAS's wimpy version of LAPACK with the real one. So far, so good. Now the setup script runs to completion, but when I try to import scipy here's what happens: % ~/bin/python Python 2.1.1 (#1, Aug 9 2001, 11:10:31) [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2 Type "copyright", "credits" or "license" for more information. line 36, in ? from scipy_base import * File "scipy_base/__init__.py", line 99, in ? import fastumath ImportError: No module named fastumath To summarize my setup, I'm using the following: Python 2.1.1 Numeric 21.0 ATLAS 3.4.1 f2py 2.23.190... gcc 2.95.4 g77 (0.5.25 20010319) on a Debian Linux box. Since I've had to go back and forth in the build process a few times, I wonder if I might be able to just do the equivalent of 'make clean' and start it fresh. I'm not very familiar with setup.py (why is this better than make?) - can anyone tell me how to do the equivalent of 'make clean'? What is 'fastumath'? Where are the sources? How do I get them to build? Thanks a lot for the tips so far - I feel like I'm close, and I'm drooling at the possibility of having LAPACK/BLAS power available from Python! -Frank At 01:00 PM 10/5/2002, you wrote:
Message: 6 From: "Travis N. Vaught" <travis@enthought.com> To: <scipy-user@scipy.net> Subject: RE: [SciPy-user] problems building - missing cvs/SciPy/xplt/gistCmodule.c Date: Fri, 4 Oct 2002 16:48:57 -0500 Reply-To: scipy-user@scipy.net
Frank,
If you're OK with skipping xplt install (and it sounds like you are), you may edit the scipy/setup.py file like so:
Line 103 old: unix_packages = ['xplt']
Line 103 new: unix_packages = []
I haven't tried this, but it should work and costs little to test.
Travis
PhD, Computational Biologist, Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA. Tel: 617-432-3555 Fax: 617-432-3557 http://llama.med.harvard.edu/~fgibbons
Frank, I had problems building from tar files but after I checked out CVS tree directly (as Travis suggested) everything worked fine (including Chaco that I installed write after finishing scipy). I am running on RedHat 7.3 though, so versions for python, gcc, etc. are a bit different. You can clean your current built by: python setup.py clean I don't know why your build wasn't successful though, so maybe someone else can help with that... Good Luck, -Viktor Frank Gibbons wrote:
Travis,
I tried this, and it does allow me to finish building, but why are there no files in the xplt directory (except for setup_xplt.py), when I can see them on the CVS page. I see that someone else had a similar problem to me, with the downloaded tarball.
Thanks for the tip. However....
I realized that I needed to get the full LAPACK installation (and therefore had to also build BLAS since I have no vendor-supplied version). My full LAPACK weighs in at a healthy 6MB, and I followed the procedures in the installation guide to replace ATLAS's wimpy version of LAPACK with the real one. So far, so good. Now the setup script runs to completion, but when I try to import scipy here's what happens:
import scipy Traceback (most recent call last): File "<stdin>", line 1, in ? File "/home/fgibbons/lib/python2.1/site-packages/scipy/__init__.py",
% ~/bin/python Python 2.1.1 (#1, Aug 9 2001, 11:10:31) [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2 Type "copyright", "credits" or "license" for more information. line 36, in ? from scipy_base import * File "scipy_base/__init__.py", line 99, in ? import fastumath ImportError: No module named fastumath
To summarize my setup, I'm using the following:
Python 2.1.1 Numeric 21.0 ATLAS 3.4.1 f2py 2.23.190... gcc 2.95.4 g77 (0.5.25 20010319) on a Debian Linux box.
Since I've had to go back and forth in the build process a few times, I wonder if I might be able to just do the equivalent of 'make clean' and start it fresh. I'm not very familiar with setup.py (why is this better than make?) - can anyone tell me how to do the equivalent of 'make clean'?
What is 'fastumath'? Where are the sources? How do I get them to build?
Thanks a lot for the tips so far - I feel like I'm close, and I'm drooling at the possibility of having LAPACK/BLAS power available from Python!
-Frank
At 01:00 PM 10/5/2002, you wrote:
Message: 6 From: "Travis N. Vaught" <travis@enthought.com> To: <scipy-user@scipy.net> Subject: RE: [SciPy-user] problems building - missing cvs/SciPy/xplt/gistCmodule.c Date: Fri, 4 Oct 2002 16:48:57 -0500 Reply-To: scipy-user@scipy.net
Frank,
If you're OK with skipping xplt install (and it sounds like you are), you may edit the scipy/setup.py file like so:
Line 103 old: unix_packages = ['xplt']
Line 103 new: unix_packages = []
I haven't tried this, but it should work and costs little to test.
Travis
PhD, Computational Biologist, Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA. Tel: 617-432-3555 Fax: 617-432-3557 http://llama.med.harvard.edu/~fgibbons
_______________________________________________ SciPy-user mailing list SciPy-user@scipy.net http://www.scipy.net/mailman/listinfo/scipy-user
Hey Frank, It sounds like you may be trying to import scipy from a python session started in your build directory. Try switching to another dir (like ~/ or /) and begin the python session and try an import. I won't tell you how many times I have to force myself to remember this each week. Travis
-----Original Message----- From: scipy-user-admin@scipy.net [mailto:scipy-user-admin@scipy.net]On Behalf Of Viktor Hornak Sent: Monday, October 07, 2002 1:58 PM To: scipy-user@scipy.net Subject: Re: [SciPy-user] Re: SciPy-user digest, Vol 1 #201 - 7 msgs
Frank,
I had problems building from tar files but after I checked out CVS tree directly (as Travis suggested) everything worked fine (including Chaco that I installed write after finishing scipy). I am running on RedHat 7.3 though, so versions for python, gcc, etc. are a bit different. You can clean your current built by: python setup.py clean I don't know why your build wasn't successful though, so maybe someone else can help with that...
Good Luck, -Viktor
Frank Gibbons wrote:
Travis,
I tried this, and it does allow me to finish building, but why are there no files in the xplt directory (except for setup_xplt.py), when I can see them on the CVS page. I see that someone else had a similar problem to me, with the downloaded tarball.
Thanks for the tip. However....
I realized that I needed to get the full LAPACK installation (and therefore had to also build BLAS since I have no vendor-supplied version). My full LAPACK weighs in at a healthy 6MB, and I followed the procedures in the installation guide to replace ATLAS's wimpy version of LAPACK with the real one. So far, so good. Now the setup script runs to completion, but when I try to import scipy here's what happens:
import scipy Traceback (most recent call last): File "<stdin>", line 1, in ? File "/home/fgibbons/lib/python2.1/site-packages/scipy/__init__.py",
% ~/bin/python Python 2.1.1 (#1, Aug 9 2001, 11:10:31) [GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2 Type "copyright", "credits" or "license" for more information. line 36, in ? from scipy_base import * File "scipy_base/__init__.py", line 99, in ? import fastumath ImportError: No module named fastumath
To summarize my setup, I'm using the following:
Python 2.1.1 Numeric 21.0 ATLAS 3.4.1 f2py 2.23.190... gcc 2.95.4 g77 (0.5.25 20010319) on a Debian Linux box.
Since I've had to go back and forth in the build process a few times, I wonder if I might be able to just do the equivalent of 'make clean' and start it fresh. I'm not very familiar with setup.py (why is this better than make?) - can anyone tell me how to do the equivalent of 'make clean'?
What is 'fastumath'? Where are the sources? How do I get them to build?
Thanks a lot for the tips so far - I feel like I'm close, and I'm drooling at the possibility of having LAPACK/BLAS power available from Python!
-Frank
At 01:00 PM 10/5/2002, you wrote:
Message: 6 From: "Travis N. Vaught" <travis@enthought.com> To: <scipy-user@scipy.net> Subject: RE: [SciPy-user] problems building - missing cvs/SciPy/xplt/gistCmodule.c Date: Fri, 4 Oct 2002 16:48:57 -0500 Reply-To: scipy-user@scipy.net
Frank,
If you're OK with skipping xplt install (and it sounds like you are), you may edit the scipy/setup.py file like so:
Line 103 old: unix_packages = ['xplt']
Line 103 new: unix_packages = []
I haven't tried this, but it should work and costs little to test.
Travis
PhD, Computational Biologist, Harvard Medical School BCMP/SGM-322, 250 Longwood Ave, Boston MA 02115, USA. Tel: 617-432-3555 Fax: 617-432-3557 http://llama.med.harvard.edu/~fgibbons
_______________________________________________ SciPy-user mailing list SciPy-user@scipy.net http://www.scipy.net/mailman/listinfo/scipy-user
_______________________________________________ SciPy-user mailing list SciPy-user@scipy.net http://www.scipy.net/mailman/listinfo/scipy-user
Travis> It sounds like you may be trying to import scipy from a python Travis> session started in your build directory. Try switching to Travis> another dir (like ~/ or /) and begin the python session and try Travis> an import. I won't tell you how many times I have to force Travis> myself to remember this each week. I get bitten by this occasionally as well. It's mildly irksome that I need to install scipy to test it. Has anyone debugged this problem? -- Skip Montanaro - skip@pobox.com "Airplanes don't fly until the paperwork equals the weight of the aircraft. Same with i18N." - from the "Perl, Unicode and i18N FAQ"
participants (4)
-
Frank Gibbons -
Skip Montanaro -
Travis N. Vaught -
Viktor Hornak