[SciPy-user] odepack or f2py error

Fabrice Silva silva at lma.cnrs-mrs.fr
Mon Jul 14 14:12:21 EDT 2008


Le samedi 12 juillet 2008 à 12:22 +0000, Pauli Virtanen a écrit :
> Sat, 12 Jul 2008 13:32:29 +0200, Fabrice Silva wrote:
> > How can I check gfortran is used by f2py ?
> Look at the output when scipy is compiled, or try to ldd scipy/integrate/
> vode.so and check if it's linked with libgfortran.

I am using scipy from debian packages, so I did not compile it. The ldd
command on either vode.so or my share object from f2py gives the
following ouput :
        ldd /usr/lib/python2.5/site-packages/scipy/integrate/vode.so
        	linux-gate.so.1 =>  (0xb7f25000)
        	libblas.so.3gf => /usr/lib/atlas/libblas.so.3gf (0xb7b83000)
        	libgfortran.so.3 => /usr/lib/libgfortran.so.3 (0xb7ad1000)
        	libm.so.6 => /lib/i686/cmov/libm.so.6 (0xb7aab000)
        	libgcc_s.so.1 => /lib/libgcc_s.so.1 (0xb7a9e000)
        	libc.so.6 => /lib/i686/cmov/libc.so.6 (0xb7943000)
        	/lib/ld-linux.so.2 (0xb7f26000)

> 	http://scipy.org/scipy/scipy/ticket/696

You may see that on this ticket, I've provided an example of small
computation where the solver reach the mxstep number of iteration, but
the error message is nicely displayed. The small example is also written
in fortran and compiled with f2py in a identical manner as I have done
with my study case...

-- 
Fabrice Silva <silva at lma.cnrs-mrs.fr>
LMA UPR CNRS 7051 - équipe S2M




More information about the SciPy-User mailing list