Re: [Numpy-discussion] Unhandled floating point exception running test in numpy-1.0.3 and svn 3875
On 6/23/07, rex [EMAIL PROTECTED] wrote: Stefan van der Walt [EMAIL PROTECTED] [2007-06-23 15:06]: On Sat, Jun 23, 2007 at 07:35:35PM +, John Ollinger wrote: I have just been updating our version of Python, numpy and scipy and have run into a floating point exception that crashes Python when I test the release. What do you mean by crash? Is anything printed? Do older versions of numpy still work? I am running gcc 3.3.1 on SuSe Linux 2.4.21-144-smp4G. The error first occurred with numpy-1.0.3. I downloaded svn 3875 when I then read the scipy web page and installed the latest subversion. The test command I am using is python -c 'import numpy; numpy.test(level=1,verbosity==2)' and occurs during the matvec test. This test uses rand to generate 10x8 and 8x1 It may be worth checking whether the new version of numpy is picked up. You can do that using import numpy as N print N.__version__ We have a build slave with a very similar setup to yours (see http://buildbot.scipy.org) and everything seems to be fine. It's somewhat different: SUSE 10.2 Core 2 Duo 32-bit Kernel 2.6.18.2-34-default gcc version 4.1.2 20061115 (prerelease) (SUSE Linux) Python 2.5 (r25:51908, Nov 27 2006 print N.__version__ 1.0.4.dev3868 python -c 'import numpy; numpy.test(level=1,verbosity=2)' [...] Ran 590 tests in 0.473s Do you use Atlas? If so, did you compile it yourself or did you use a package? There is a bug in some older 64 bit Atlas packages running on newer intel hardware that generates illegal instruction exceptions and I am wondering if you may have found a new 32 bit bug. One way to check this is to multiply two big matrices together. There are many paths through Atlas, so the known bug is not encountered in all matrix multiplications, and perhaps not for all floating values either. Chuck ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] [ANN]New numpy, scipy and atlas rpms for FC 5, 6 and 7 and openSUSE (with 64 bits arch support)
Hi there, After quite some pain, I finally managed to build a LAPACK + ATLAS rpm useful for numpy and scipy. Read the following if you use Fedora Core or OpenSuse and are tired to install unsuccessfully numpy, scipy, BLAS, LAPACK or ATLAS. Instructions are given there: http://www.scipy.org/Installing_SciPy/Linux (ashigabou repository) Basically: - Fedora Core 5, 6 and 7 and openSUSE 10.2 are supported (x86, and x86_64 for FC 7 and openSuse). - binary rpms for numpy, scipy and blas/lapack dependencies. - source rpm for atlas, for a really easy, 3 commands build of ATLAS (should work for both x86 and x86_64). numpy and scipy are the last releases, including some backported changes to make it work on 64 bits. Atlas is the last developement version, with a trivial patch to build shared blas and lapack which can be used as drop in replacements for netlib blas and lapack. I would like to hear people complains. If people want other distributions supported by the opensuse build system (such as mandriva), I would like to hear it too. cheers, David ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] Unhandled floating point exception running test in numpy-1.0.3 and svn 3875
Charles R Harris [EMAIL PROTECTED] [2007-06-24 06:22]: On 6/23/07, rex [EMAIL PROTECTED] wrote: Stefan van der Walt [EMAIL PROTECTED] [2007-06-23 15:06]: On Sat, Jun 23, 2007 at 07:35:35PM +, John Ollinger wrote: I have just been updating our version of Python, numpy and scipy and have run into a floating point exception that crashes Python when I test the release. What do you mean by crash? Is anything printed? Do older versions of numpy still work? John needs to respond to this. I am running gcc 3.3.1 on SuSe Linux 2.4.21-144-smp4G. The error first occurred with numpy-1.0.3. I downloaded svn 3875 when I then read the scipy web page and installed the latest subversion. The test command I am using is python -c 'import numpy; numpy.test(level=1,verbosity==2)' and occurs during the matvec test. This test uses rand to generate 10x8 and 8x1 It may be worth checking whether the new version of numpy is picked up. You can do that using import numpy as N print N.__version__ We have a build slave with a very similar setup to yours (see http://buildbot.scipy.org ) and everything seems to be fine. It's somewhat different: SUSE 10.2 Core 2 Duo 32-bit Kernel 2.6.18.2-34-default gcc version 4.1.2 20061115 (prerelease) (SUSE Linux) Python 2.5 (r25:51908, Nov 27 2006 print N.__version__ 1.0.4.dev3868 python -c 'import numpy; numpy.test(level=1,verbosity=2)' [...] Ran 590 tests in 0.473s Do you use Atlas? If so, did you compile it yourself or did you use a package? There is a bug in some older 64 bit Atlas packages running on newer intel hardware that generates illegal instruction exceptions and I am wondering if you may have found a new 32 bit bug. One way to check this is to multiply two big matrices together. There are many paths through Atlas, so the known bug is not encountered in all matrix multiplications, and perhaps not for all floating values either. The above system is running the http://buildbot.scipy.org/ buildbot with no errors. It doesn't appear to build ATLAS. It's John's older system (gcc 3.3.1 2.4.21-144-smp4G) -- AFAIK, 2.4.21 was used in SUSE 9.0. Current version is 10.2 -- that is throwing an error, not mine. -rex ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] Help installing numpy 1.0.2 on LINUX
I am trying to install nunpy 1.0.2 on aMandrake Linux system kernel 2.6.17mdk. I have followed the Install instructions in numpy. I am including a small portion of the python setup.py install command: Running from numpy source directory. F2PY Version 2_3649 blas_opt_info: blas_mkl_info: libraries mkl,vml,guide not found in /usr/local/lib/atlas NOT AVAILABLE atlas_blas_threads_info: Setting PTATLAS=ATLAS libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib/atlas NOT AVAILABLE atlas_blas_info: FOUND: libraries = ['f77blas', 'cblas', 'atlas'] library_dirs = ['/usr/local/lib/atlas'] language = c Could not locate executable g77 Could not locate executable f95 customize GnuFCompiler customize GnuFCompiler customize GnuFCompiler using config compiling '_configtest.c': /* This file is generated from numpy_distutils/system_info.py */ void ATL_buildinfo(void); int main(void) { ATL_buildinfo(); return 0; } C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC compile options: '-c' gcc: _configtest.c gcc -pthread _configtest.o -L/usr/local/lib/atlas -lf77blas -lcblas -latlas -o _configtest ATLAS version 3.6.0 built by root on Fri Jun 15 18:59:04 EDT 2007: UNAME: Linux localhost 2.6.17-5mdv #1 SMP Wed Sep 13 14:32:31 EDT 2006 i686 Celeron (Mendocino) GNU/Linux INSTFLG : MMDEF: ARCHDEF : F2CDEFS : -DAdd__ -DStringSunStyle CACHEEDGE: 131072 F77 : /usr/local//gcc2.95.3/bin/g77, version GNU Fortran 0.5.25 20010315 (release) F77FLAGS : -fomit-frame-pointer -O CC : /usr/local/gcc2.95.3/bin/gcc, version 2.95.3 CC FLAGS : -fomit-frame-pointer -O3 -funroll-all-loops MCC : /usr/local/gcc2.95.3/bin/gcc, version 2.95.3 MCCFLAGS : -fomit-frame-pointer -O success! removing: _configtest.c _configtest.o _configtest FOUND: libraries = ['f77blas', 'cblas', 'atlas'] library_dirs = ['/usr/local/lib/atlas'] language = c define_macros = [('ATLAS_INFO', '\\3.6.0\\')] lapack_opt_info: lapack_mkl_info: mkl_info: libraries mkl,vml,guide not found in /usr/local/lib/atlas NOT AVAILABLE NOT AVAILABLE atlas_threads_info: Setting PTATLAS=ATLAS libraries ptf77blas,ptcblas,atlas not found in /usr/local/lib/atlas libraries lapack_atlas not found in /usr/local/lib/atlas numpy.distutils.system_info.atlas_threads_info NOT AVAILABLE atlas_info: libraries lapack_atlas not found in /usr/local/lib/atlas numpy.distutils.system_info.atlas_info FOUND: libraries = ['lapack', 'f77blas', 'cblas', 'atlas'] library_dirs = ['/usr/local/lib/atlas'] language = f77 customize GnuFCompiler customize GnuFCompiler customize GnuFCompiler using config compiling '_configtest.c': /* This file is generated from numpy_distutils/system_info.py */ void ATL_buildinfo(void); int main(void) { ATL_buildinfo(); return 0; } C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC compile options: '-c' gcc: _configtest.c gcc -pthread _configtest.o -L/usr/local/lib/atlas -llapack -lf77blas -lcblas -latlas -o _configtest ATLAS version 3.6.0 built by root on Fri Jun 15 18:59:04 EDT 2007: UNAME: Linux localhost 2.6.17-5mdv #1 SMP Wed Sep 13 14:32:31 EDT 2006 i686 Celeron (Mendocino) GNU/Linux INSTFLG : MMDEF: ARCHDEF : F2CDEFS : -DAdd__ -DStringSunStyle CACHEEDGE: 131072 F77 : /usr/local//gcc2.95.3/bin/g77, version GNU Fortran 0.5.25 20010315 (release) F77FLAGS : -fomit-frame-pointer -O CC : /usr/local/gcc2.95.3/bin/gcc, version 2.95.3 CC FLAGS : -fomit-frame-pointer -O3 -funroll-all-loops MCC : /usr/local/gcc2.95.3/bin/gcc, version 2.95.3 MCCFLAGS : -fomit-frame-pointer -O success! removing: _configtest.c _configtest.o _configtest FOUND: libraries = ['lapack', 'f77blas', 'cblas', 'atlas'] library_dirs = ['/usr/local/lib/atlas'] language = f77 define_macros = [('ATLAS_INFO', '\\3.6.0\\')] running install running build running config_fc running build_src building py_modules sources creating build/src.linux-i686-2.5 creating build/src.linux-i686-2.5/numpy creating build/src.linux-i686-2.5/numpy/distutils building extension numpy.core.multiarray sources creating build/src.linux-i686-2.5/numpy/core Generating build/src.linux-i686-2.5/numpy/core/config.h customize GnuFCompiler customize GnuFCompiler customize GnuFCompiler using config C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC When I try to run numpy.test(level=1) I get: import numpy numpy.test(level=1) Traceback (most recent call last) File stdln. line 1, in module AttributeError: 'module' has no attribute 'test' Thank you for your help or suggestions. John -- ___ Surf the Web in a faster, safer and easier way: Download Opera 9 at http://www.opera.com
Re: [Numpy-discussion] Help installing numpy 1.0.2 on LINUX
On Sun, Jun 24, 2007 at 05:58:33PM +0100, John Pruce wrote: When I try to run numpy.test(level=1) I get: import numpy numpy.test(level=1) Traceback (most recent call last) File stdln. line 1, in module AttributeError: 'module' has no attribute 'test' Thank you for your help or suggestions. Are you running from the numpy source directory? If so, change out of it and try again. Stéfan ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] Help installing numpy 1.0.2 on LINUX
- Original Message - From: Stefan van der Walt [EMAIL PROTECTED] To: numpy-discussion@scipy.org Subject: Re: [Numpy-discussion] Help installing numpy 1.0.2 on LINUX Date: Sun, 24 Jun 2007 19:48:14 +0200 On Sun, Jun 24, 2007 at 05:58:33PM +0100, John Pruce wrote: When I try to run numpy.test(level=1) I get: import numpy numpy.test(level=1) Traceback (most recent call last) File stdln. line 1, in module AttributeError: 'module' has no attribute 'test' Thank you for your help or suggestions. Are you running from the numpy source directory? If so, change out of it and try again. Stéfan ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion Thank you That was the problem John -- ___ Surf the Web in a faster, safer and easier way: Download Opera 9 at http://www.opera.com Powered by Outblaze ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] Unhandled floating point exception running test in numpy-1.0.3 and svn 3875
At 11:09 AM 6/24/2007, you wrote: On 6/24/07, rex [EMAIL PROTECTED] wrote: Charles R Harris [EMAIL PROTECTED] [2007-06-24 06:22]: On 6/23/07, rex [EMAIL PROTECTED] wrote: Stefan van der Walt [EMAIL PROTECTED] [2007-06-23 15:06]: On Sat, Jun 23, 2007 at 07:35:35PM +, John Ollinger wrote: I have just been updating our version of Python, numpy and scipy and have run into a floating point exception that crashes Python when I test the release. What do you mean by crash? Is anything printed? Do older versions of numpy still work? John needs to respond to this. I am running gcc 3.3.1 on SuSe Linux 2.4.21-144-smp4G. The error first occurred with numpy-1.0.3. I downloaded svn 3875 when I then read the scipy web page and installed the latest subversion. The test command I am using is python -c 'import numpy; numpy.test(level=1,verbosity==2)' and occurs during the matvec test. This test uses rand to generate 10x8 and 8x1 It may be worth checking whether the new version of numpy is picked up. You can do that using import numpy as N print N.__version__ We have a build slave with a very similar setup to yours (see http://buildbot.scipy.org ) and everything seems to be fine. It's somewhat different: SUSE 10.2 Core 2 Duo 32-bit Kernel 2.6.18.2-34-default gcc version 4.1.2 20061115 (prerelease) (SUSE Linux) Python 2.5 (r25:51908, Nov 27 2006 print N.__version__ 1.0.4.dev3868 python -c 'import numpy; numpy.test(level=1,verbosity=2)' [...] Ran 590 tests in 0.473s Do you use Atlas? If so, did you compile it yourself or did you use a package? There is a bug in some older 64 bit Atlas packages running on newer intel hardware that generates illegal instruction exceptions and I am wondering if you may have found a new 32 bit bug. One way to check this is to multiply two big matrices together. There are many paths through Atlas, so the known bug is not encountered in all matrix multiplications, and perhaps not for all floating values either. The above system is running the http://buildbot.scipy.org/ buildbot with no errors. It doesn't appear to build ATLAS. It's John's older system (gcc 3.3.1 2.4.21-144-smp4G) -- AFAIK, 2.4.21 was used in SUSE 9.0. Current version is 10.2 -- that is throwing an error, not mine. Sorry, my mistake. And I misread 2.4.21 as 2.6.21, a very recent kernel. Just shows how fast the years pass by. Chuck ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion You hit the nail on the head. It is apparently an Atlas problem. I originally built numpy with optimized Atlas libraries. When I deleted these from the site.cfg file the problem went away. I apologize for not including the error message in my initial post. Here it is in case anyone cares: check_matvec (numpy.core.tests.test_numeric.test_dot)Floating point exception I appreciate the helpful responses. I will fool around with Atlas to see if I can eliminate the problem. I have to do stuff that makes people want to pay for part of today, so I might not post again for a day or two. Thanks again, John ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] Unhandled floating point exception running test in numpy-1.0.3 and svn 3875
John Ollinger [EMAIL PROTECTED] [2007-06-24 13:13]: I am running gcc 3.3.1 on SuSe Linux 2.4.21-144-smp4G. The error first occurred with numpy-1.0.3. I downloaded svn 3875 when I then read the scipy web page and installed the latest subversion. The test command I am using is python -c 'import numpy; numpy.test(level=1,verbosity==2)' and occurs during the matvec test. This test uses rand to generate 10x8 and 8x1 You hit the nail on the head. It is apparently an Atlas problem. I originally built numpy with optimized Atlas libraries. When I deleted these from the site.cfg file the problem went away. I apologize for not including the error message in my initial post. Here it is in case anyone cares: check_matvec (numpy.core.tests.test_numeric.test_dot)Floating point exception I recently built (using David Cournapeau's garnumpy) numpy and scipy using ATLAS 3.7.33, gfortran, gcc 4.1.2, SUSE 10.2 (x86 kernel 2.6.18.2-34-default), and python 2.5: python Python 2.5 (r25:51908, Nov 27 2006, 19:14:46) [GCC 4.1.2 20061115 (prerelease) (SUSE Linux)] on linux2 Type help, copyright, credits or license for more information. import numpy import scipy print numpy.__version__ 1.0.1 print scipy.__version__ 0.5.2 numpy.test(level=1,verbosity=2) [...] Ran 526 tests in 0.432s OK scipy.test(level=1,verbosity=2) [...] Ran 1596 tests in 2.942s OK There doesn't appear to be a problem with recent versions of the software. In particular, ATLAS 3.7.33 does not cause an error. Is there some reason for you to use such old software? (gcc 3.3.1 kernel 2.4.21)? What platform are you building for? -rex ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] PCA - Principal Component Analysis
I'm doing some projects in Python (system GNU / Linux - Ubuntu 7.0) about image processing. I'm needing a implementation of PCA, prefer to library for apt-get. Thanks. Alex. Novo Yahoo! Cadê? - Experimente uma nova busca. http://yahoo.com.br/oqueeuganhocomisso ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
[Numpy-discussion] qr decomposition with column pivoting/qr decomposition with householder reflections
I'm in the process of trying to convert some Matlab code into Python. There's a statement of the form: [q,r,e] = qr(A) which performs a qr-decomposition of A, but then also returns a 'permutation' matrix. The purpose of this is to ensure that the values along r's diagonal are decreasing. I believe this technique is called qr decomposition with column pivoting or (equivalently) qr decomposition with householder reflections. I have not been able to find an implementation of this within numpy. Does one exist? Or should I come to truly understand this algorithm (prob'ly a good idea regardless) and implement it? Thanks, Steven ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] qr decomposition with column pivoting/qr decomposition with householder reflections
On 6/21/07, traveller3141 [EMAIL PROTECTED] wrote: I'm in the process of trying to convert some Matlab code into Python. There's a statement of the form: [q,r,e] = qr(A) which performs a qr-decomposition of A, but then also returns a 'permutation' matrix. The purpose of this is to ensure that the values along r's diagonal are decreasing. I believe this technique is called qr decomposition with column pivoting or (equivalently) qr decomposition with householder reflections. There is a qr version in numpy, numpy.linalg.qr, but it only returns the factors q and r. The underlying lapack routines are {dz}geqrf and {dz}orgqr, the latter converting the product of Householder reflections into the orthogonal matrix q. Column pivoting is not used in {dz}geqrf, but it *is* used in {dz}geqpf. The versions with column pivoting are probably more accurate and also allow fixing certain columns to the front of the array, a useful thing in some cases, so I don't know why we chose the first rather than the second. I suspect the decision was made in Numeric long ago and the simplest function was chosen. The column pivoting version isn't in scipy either and it probably should be. If you need it, it shouldn't be hard to add. Chuck ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] PCA - Principal Component Analysis
IMO the Modular toolkit for Data Processing (MDP) has a fairly good and straightforward PCA implementation, among other good tools: mdp-toolkit.sourceforge.net/ I have no idea what apt-get is, though, so I don't know if this will be helpful or not! -Rob On 21/06/07, Alex Torquato S. Carneiro [EMAIL PROTECTED] wrote: I'm doing some projects in Python (system GNU / Linux - Ubuntu 7.0) about image processing. I'm needing a implementation of PCA, prefer to library for apt-get. Thanks. Alex. Novo Yahoo! Cadê? - Experimente uma nova busca. ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
Re: [Numpy-discussion] PCA - Principal Component Analysis
On 6/24/07, Rob Clewley [EMAIL PROTECTED] wrote: IMO the Modular toolkit for Data Processing (MDP) has a fairly good and straightforward PCA implementation, among other good tools: mdp-toolkit.sourceforge.net/ I have no idea what apt-get is, though, so I don't know if this will be helpful or not! -Rob Apt-get fetches and installs packages for Debian and Debian based Linux distributions like Ubuntu. Chuck ___ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion