Mailman 3 November 2008 - NumPy-Discussion

extension module with swig
by Giovanni Samaey 04 Nov '08

04 Nov '08

Dear all, I am unsure about the correct place to put this question -- If this isn't the correct list, please let me know which place is more appropriate. I am trying to build an extension module in python that calls a C routine that depends on the GNU Scientific Library. I am using swig to wrap the c-file and the numpy.i file that is included with numpy for the typemaps. To build the module, I am using distutils, with a modified version of the setup.py file that was included with the examples (in order to link against gsl). Although the build succeeds wtihout problems, I am getting "undefined symbol" errors from gsl when importing the module in python. I am attaching the source code of the simplest possible example below together with the error messages. Note that I am trying to link against my own gsl in my home directory. There is no system-wide one available. I get the feeling that I am doing something silly, but I am unable to put the finger on it. Any help would be greatly appreciated. As a c-file I have : test.c #include <gsl/gsl_rng.h> #include <gsl/gsl_randist.h> # include <math.h> double var; void test(double *x, double *y, double *z, int n) { int i; const gsl_rng_type *T; gsl_rng *r; // gsl_rng_env_setup(); // T = gsl_rng_default; T = gsl_rng_mt19937; r = gsl_rng_alloc(T); for (i=0;i<n;i++) { *(z++)=x[i]+y[i]+var; } var += gsl_rng_uniform(r); } As a header file : test.h extern double var; void test(double *x, double *y, double *z, int n); The swig interface file: test.i /* test.i */ %module test %{ #define SWIG_FILE_WITH_INIT %} %include "numpy.i" %init %{ import_array(); %} %inline %{ #include "test.h" %} %apply (double* IN_ARRAY1, int DIM1) {(double* x, int n1), (double* y, int n2)} %apply (double* ARGOUT_ARRAY1, int DIM1) {(double* z, int m)}; %rename (test) my_test; %inline %{ void my_test(double* x, int n1, double* y, int n2, double* z, int m) { test(x,y,z,m); } %} %ignore test; %include "test.h" The setup.py file is : #! /usr/bin/env python # System imports from distutils.core import * from distutils import sysconfig # Third-party modules - we depend on numpy for everything import numpy # Obtain the numpy include directory. This logic works across numpy versions.try: numpy_include = numpy.get_include() except AttributeError: numpy_include = numpy.get_numpy_include() # dot extension module _test = Extension("_test", ["test_wrap.c", "test.c"], include_dirs = [numpy_include,'/data/home/ u0038151/include'], library_dirs = ['/data/home/u0038151/lib'] ) # NumyTypemapTests setup setup(name = ["test"], description = "test c module swigging", author = "Giovanni Samaey", py_modules = ["test"], ext_modules = [_test ] ) The error message is the following: 0 u0038151@lo-03-02 dot2 $ python -c "import test" Traceback (most recent call last): File "<string>", line 1, in ? File "test.py", line 7, in ? import _test ImportError: ./_test.so: undefined symbol: gsl_rng_mt19937

4 6

Re: [Numpy-discussion] atlas not found, why?
by Scott Sinclair 04 Nov '08

04 Nov '08

>> "T J" <tjhnson(a)gmail.com> 11/04/08 12:59 AM On Mon, Nov 3, 2008 at 10:46 AM, T J <tjhnson(a)gmail.com> wrote: > > Since these are all in the standard locations, I am building without a > site.cfg. Here is the beginning info: > >> Apparently, this is not enough. Only if I also set the ATLAS >> environment variable am I able to get this working as expected. >> So can someone explain why I *must* define ATLAS. I tried a number of >> variations on site.cfg and could not get numpy to find atlas with any >> of them. When I upgraded from Ubuntu 8.04 to 8.10 it broke my Numpy install (mighty irritating). I haven't looked in detail, but It seems as if some of the packages related to ATLAS have been renamed and the post-upgrade "cleanup" removed the development packages as it considered them obsolete. I made sure that libatlas-base-dev and libatlas-sse2-dev (this machine supports sse2) were installed and was able to rebuild and install Numpy with no problems. I do not have an ATLAS environment variable. Cheers, Scott Please find our Email Disclaimer here: http://www.ukzn.ac.za/disclaimer/

2 1

Compile error during 32 bit compile on 64 bit machine
by Brian 04 Nov '08

04 Nov '08

Greeting, I am trying to install numpy on the same system into to different prefixes, one dedicated to 32 bit stuff and the other to 64 bit. The machine is a dual CPU Core 2 Duo machine, that is a 64 bit native machine. Getting the 64 bit version to work is no problem. But cross compiling numpy for 32 bit binaries is very troublesome. I have compiled all the dependences using "-m32" including Python itself. All my other packages compile using the -m32 flag without problems. I have set CFLAGS, LDFLAGS, BASECFLAGS, CPPFLAGS, and CXXFLAGS all to reflect the change in bits to use. Please find below the error I receive when compiling with: python setup.py build --fcompiler=gnu95. The error appears to be because it is trying to link _configtest.o without the -m32 flag and thinks it is making a 64 bit binary. If I edit line 378 of ccompiler.py to add "-m32" to the link_opts by hand I can get past this error to a new one which indicates that main was never declared in the linalg package. If that is the more useful starting point I can post those errors as well. Any help would be greatly appreciated. Regards, Brian ############### ERROR BELOW ##################### C compiler: gcc -pthread -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict- prototypes -m32 -m32 -fPIC compile options: '-I/rh5stuff/32bit/include/python2.6 -Inumpy/core/src -Inumpy/core/include -I/rh5stuff/32bit/include/python2.6 -c' gcc: _configtest.c gcc -pthread _configtest.o -L/rh5stuff/32bit/lib -L/usr/local/lib -L/ usr/lib -o _configtest /usr/bin/ld: skipping incompatible /usr/lib/libpthread.so when searching for -lpthread /usr/bin/ld: skipping incompatible /usr/lib/libpthread.a when searching for -lpthread /usr/bin/ld: skipping incompatible /usr/lib/libc.so when searching for -lc /usr/bin/ld: skipping incompatible /usr/lib/libc.a when searching for - lc /usr/bin/ld: warning: i386 architecture of input file `_configtest.o' is incompatible with i386:x86-64 output _configtest failure. removing: _configtest.c _configtest.o _configtest

2 1

computing average distance
by Paul Rudin 03 Nov '08

03 Nov '08

I'm experimenting with numpy and I've just written the code below, which computes the thing I want (I think). Self.bits is an RxRxR array representing a voxelized 3d model - values are either 0 or 1. I can't help thinking that's there must be a much nicer way to do it. Any suggestions? centre = numpy.array(scipy.ndimage.measurements.center_of_mass(self.bits)) vectors = [] for x in xrange(R): for y in xrange(R): for z in xrange(R): if self.bits[x,y,z]: vectors.append([x,y,z]) vectors = numpy.array(vectors) distances = numpy.sqrt(numpy.sum((vectors-centre) ** 2.0, axis=1)) av_dist = numpy.average(distances)

4 4

fink python26 and numpy 1.2.1
by Gideon Simpson 03 Nov '08

03 Nov '08

Not sure if this is an issue with numpy or an issue with fink python 2.6, but when trying to build numpy, I get the following error: gcc -L/sw/lib -bundle /sw/lib/python2.6/config -lpython2.6 build/ temp.macosx-10.5-i386-2.6/numpy/core/src/multiarraymodule.o -o build/ lib.macosx-10.5-i386-2.6/numpy/core/multiarray.so ld: library not found for -lpython2.6 collect2: ld returned 1 exit status ld: library not found for -lpython2.6 collect2: ld returned 1 exit status error: Command "gcc -L/sw/lib -bundle /sw/lib/python2.6/config - lpython2.6 build/temp.macosx-10.5-i386-2.6/numpy/core/src/ multiarraymodule.o -o build/lib.macosx-10.5-i386-2.6/numpy/core/ multiarray.so" failed with exit status 1 -gideon

3 3

Simplifying compiler optimization flags logic (fortran compilers)
by David Cournapeau 03 Nov '08

03 Nov '08

Hi, I was wondering whether it was really worth having a lot of magic going on in fcompilers for flags like -msse2 and co (everything done in get_flags_arch, for example). It is quite fragile (we had several problems wrt buggy compilers, buggy CPU detection), and I am not sure it buys us much anyway. Did some people notice a difference between gfortran -O3 -msse2 and gfortran -O3 ? thanks, David

5 4

numpy function to compute sample ranks
by Abhimanyu Lad 02 Nov '08

02 Nov '08

Hi, Is there a direct or indirect way in numpy to compute the sample ranks of a given array, i.e. the equivalent of rank() in R. I am looking for: rank(array([6,8,4,1,9])) -> array([2,3,1,0,4]) Is there some clever use of argsort() that I am missing? Thanks, Abhi

3 2

Re: [Numpy-discussion] Installation Trouble
by T J 02 Nov '08

02 Nov '08

Sorry....wrong list. On Sun, Nov 2, 2008 at 11:34 AM, T J <tjhnson(a)gmail.com> wrote: > Hi, > > I'm having trouble installing PyUblas 0.93.1 (same problems from the > current git repository). I'm in ubuntu 8.04 with standard boost > packages (1.34.1, I believe). Do you have any suggestions? > > Thanks! >

1 0

Re: [Numpy-discussion] Complete LAPACK needed (Frank Lagor)
by Frank Lagor 01 Nov '08

01 Nov '08

Hi Everyone, I am trying to give an update here so that maybe it will be easier for someone to help me out now. I have gone through the lapack installation and the atlas installation and everything seems to have worked fine. Everything was compiled using the 'fPIC' option and I figured out how to make shared blas and lapack library files with atlas. Also, I figured out how to tell Atlas that I want a COMPLETE lapack. Now I am having trouble with the numpy install. In the site.cfg file, I included the location where I have the include and lib files for lapack. After install, when I test, I get the following: [dfranci@alpha ~]$ python -c 'import numpy; numpy.test()' Traceback (most recent call last): File "<string>", line 1, in <module> File "/ronaldo/pkg/dfranci/Python-2.5.2/lib/python2.5/site-packages/numpy/__init__.py", line 143, in <module> import linalg File "/ronaldo/pkg/dfranci/Python-2.5.2/lib/python2.5/site-packages/numpy/linalg/__init__.py", line 47, in <module> from linalg import * File "/ronaldo/pkg/dfranci/Python-2.5.2/lib/python2.5/site-packages/numpy/linalg/linalg.py", line 29, in <module> from numpy.linalg import lapack_lite ImportError: liblapack.so: cannot open shared object file: No such file or directory I notice one thing here that concerns me: numpy is trying to import a module that it calls "lapack_lite". Does numpy always build such a module? If I am trying to get the complete lapack, did I do something wrong? Thanks in advance, Frank

3 3

installation problem on Red Hat
by Christoph Göbl 01 Nov '08

01 Nov '08

dear members, I'm very sorry to bother you with a (hopefully) simple problem... I need pyhton and the numerical package to run another program. I installed Python, it works fine. But I can't install the numpy package. To install the oder Numeric package was no problem, but I need the newer numpy... after "python setup.py install" I get an error message after sime while: ... compiling C sources C compiler: gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC compile options: '-DNO_ATLAS_INFO=1 -Inumpy/core/include -Ibuild/src.linux-x86_64-2.5/numpy/core/include/numpy -Inumpy/core/src -Inumpy/core/include -I/usr/local/include/python2.5 -c' /usr/local/bin/g77 -g -Wall -g -Wall -shared build/temp.linux-x86_64-2.5/numpy/linalg/lapack_litemodule.o build/temp.linux-x86_64-2.5/numpy/linalg/python_xerbla.o -L/usr/lib -llapack -lblas -lg2c -o build/lib.linux-x86_64-2.5/numpy/linalg/lapack_lite.so /usr/bin/ld: skipping incompatible /usr/lib/liblapack.so when searching for -llapack /usr/bin/ld: skipping incompatible /usr/lib/liblapack.a when searching for -llapack /usr/bin/ld: skipping incompatible /usr/lib/libblas.so when searching for -lblas /usr/bin/ld: skipping incompatible /usr/lib/libblas.a when searching for -lblas /usr/bin/ld: skipping incompatible /usr/lib/libg2c.so when searching for -lg2c /usr/bin/ld: skipping incompatible /usr/lib/libg2c.a when searching for -lg2c /usr/bin/ld: skipping incompatible /usr/lib/libfrtbegin.a when searching for -lfrtbegin /usr/bin/ld: skipping incompatible /usr/lib/libg2c.so when searching for -lg2c /usr/bin/ld: skipping incompatible /usr/lib/libg2c.a when searching for -lg2c /usr/bin/ld: cannot find -lgcc_s collect2: ld returned 1 exit status /usr/bin/ld: skipping incompatible /usr/lib/liblapack.so when searching for -llapack /usr/bin/ld: skipping incompatible /usr/lib/liblapack.a when searching for -llapack /usr/bin/ld: skipping incompatible /usr/lib/libblas.so when searching for -lblas /usr/bin/ld: skipping incompatible /usr/lib/libblas.a when searching for -lblas /usr/bin/ld: skipping incompatible /usr/lib/libg2c.so when searching for -lg2c /usr/bin/ld: skipping incompatible /usr/lib/libg2c.a when searching for -lg2c /usr/bin/ld: skipping incompatible /usr/lib/libfrtbegin.a when searching for -lfrtbegin /usr/bin/ld: skipping incompatible /usr/lib/libg2c.so when searching for -lg2c /usr/bin/ld: skipping incompatible /usr/lib/libg2c.a when searching for -lg2c /usr/bin/ld: cannot find -lgcc_s collect2: ld returned 1 exit status error: Command "/usr/local/bin/g77 -g -Wall -g -Wall -shared build/temp.linux-x86_64-2.5/numpy/linalg/lapack_litemodule.o build/temp.linux-x86_64-2.5/numpy/linalg/python_xerbla.o -L/usr/lib -llapack -lblas -lg2c -o build/lib.linux-x86_64-2.5/numpy/linalg/lapack_lite.so" failed with exit status 1 ... I don't know, what I can do... On a Suse 10.2 it ran easily, but on the other computer, Red Hat 3.4.3_9, X86_64, gcc 3.4.3 there is always this error message. I read in another forum, that a person solved a similar problem using unsetenv ldflags But - sorry I'm a Newbie in Linux and Python - there the installation was on another platform I think. Anyway, maybe it's a linking problem? thank you very much for any thoughts you may waste on my problems... best regards, Christoph _________________________________________________________________ Hol dir 30 kostenlose Emoticons für deinen Windows Live Messenger http://www.livemessenger-emoticons.com/funfamily/de-at/

7 8