Just for future reference, and because it wasn't immediately obvious from googling: the umfpack headers in Ubuntu 13.04 are in the package libsuitesparse-dev. But now the scipy build fails with some weird error. Hmmm. building extension "scipy.sparse.linalg.dsolve.umfpack.__umfpack" sources creating build/src.linux-x86_64-2.7/scipy/sparse/linalg/dsolve creating build/src.linux-x86_64-2.7/scipy/sparse/linalg/dsolve/umfpack adding 'scipy/sparse/linalg/dsolve/umfpack/umfpack.i' to sources. swig: scipy/sparse/linalg/dsolve/umfpack/umfpack.i swig -python -I/usr/include/suitesparse -I/usr/include/suitesparse -I/usr/include/suitesparse -o build/src.linux-x86_64-2.7/scipy/sparse/linalg/dsolve/umfpack/_umfpack_wrap.c -outdir build/src.linux-x86_64-2.7/scipy/sparse/linalg/dsolve/umfpack scipy/sparse/linalg/dsolve/umfpack/umfpack.i unable to execute swig: No such file or directory error: command 'swig' failed with exit status 1 On Fri, Oct 4, 2013 at 2:39 PM, Juan Nunez-Iglesias <jni.soma@gmail.com>wrote:

Thanks Josh for the detailed response and discussion! I think you're absolutely right that it should be suppressed if pyamg is available. And it's probably worth getting that in before the release. Happy to tackle this.

Juan.

PS: I'm mostly on OSX, with some Linux on the side. =)

On Fri, Oct 4, 2013 at 9:17 AM, Josh Warner <silvertrumpet999@gmail.com>wrote:

...

Final note - umfpack enables mode='cg' in random_walker, which is better than brute force but not the fastest or most efficient mode. The multi-grid preconditioned conjugate gradient method is: mode='cg_mg'. This is enabled by pyamg, which is independent of umfpack and is considerably easier to install and use than the method I just described.

Come to think of it, we probably should suppress that warning if import pyamg works. It incorrectly implies the user doesn't have the fastest mode available.

On Thursday, October 3, 2013 1:21:43 AM UTC-5, Juan Nunez-Iglesias wrote:

Hi guys,

...

How do I build scipy with umfpack to silence this error:

Exception AttributeError: "'UmfpackContext' object has no attribute '_symbolic'" in <bound method UmfpackContext.__del__ of <scipy.sparse.linalg.dsolve.**umfpack.umfpack.UmfpackContext object at 0x101b62050>> ignored

and speed up random_walker? I've just been doing a pip install scipy till now.

Juan.

-- You received this message because you are subscribed to the Google Groups "scikit-image" group. To unsubscribe from this group and stop receiving emails from it, send an email to scikit-image+unsubscribe@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.