[SciPy-user] ... object has no attribute 'zgelss'

H Jansen h.jansen at fel.tno.nl
Fri Feb 13 16:58:47 CST 2004


On Fri, 2004-02-13 at 16:23, Pearu Peterson wrote:
> On Fri, 13 Feb 2004, H Jansen wrote:
> 
> > Many Lapack routines can't be found when I do:
> > 
> > import scipy
> > scipy.test(level=1)
> > 
> > For instance:
> <snip> 
> > ... << more of the same, all related to lapack subroutines >>
> > 
> > When I do "nm liblapack.a | grep zgelss" code for "zgelss_";
> 
> ??? So what is the result then? Are zgelss_ and others defined or not?
> 
> > likewise 
> > with all other routines. I'm using ATLAS-3.6.0 and my liblapack is 
> > complete (appr. 6.8 Mb, routines added from CLAPACK).
> 
> CLAPACK??? Unless CLAPACK provides symbol names with '_' suffix, you 
> should complete ATLAS liblapack.a with Fortran LAPACK.

With Fortran Lapack everything now works fine!

Thanks! Henk.


> > Apparently,
> > there's an interface mismatch between ATLAS and scipy 
> > (SciPy-0.2.0_alpha_200.4161).
> > 
> > The scipy source builts with the command
> > 
> > 	python setup.py build build_ext `pkg-config atlas --cflags --libs`
> > 
> > resulting the linalg shared modules to be linked as
> > 
> > 	-L/opt/src/math/numerical/matrix/atlas/current/lib/Linux_P4SSE2 
> > -llapack -lf77blas -lcblas -latlas -lg2c
> 
> I suspect that scipy setup finds different atlas/lapack from you system 
> than that you specify with `pkd-config`. Try building scipy as follows
> 
>   rm -rf build
>   export ATLAS=/opt/src/math/numerical/matrix/atlas/current/lib/Linux_P4SSE2
>   python setup.py build
> 
> HTH,
> Pearu
> 
> _______________________________________________
> SciPy-user mailing list
> SciPy-user at scipy.net
> http://www.scipy.net/mailman/listinfo/scipy-user
-- 
Henk Jansen <h.jansen at fel.tno.nl>
TNO Physics and Electronics Laboratory


-- 
------------------------------------------------------------------------------
The disclaimer that applies to e-mail from
TNO Physics and Electronics Laboratory
can be found on: http://www.tno.nl/disclaimer/email.html
------------------------------------------------------------------------------



More information about the SciPy-user mailing list