[SciPy-dev] problem with linalg.cholesky?
Nils Wagner
nwagner at mecha.uni-stuttgart.de
Mon Nov 28 15:25:08 EST 2005
On Mon, 28 Nov 2005 19:37:53 +0000
Andrew Jaffe <a.h.jaffe at gmail.com> wrote:
> hi all,
>
> (apologies that a similar question has appeared
>elsewhere...)
>
> In the newest incarnation of scipy_core, I am having
>trouble with the
> cholesky(a) routine. Here is some minimal code
>reproducing the bug
> (on OS X)
>
> ------------------------------------------------------------
> from scipy import identity, __core_version__, Float64
> import scipy.linalg as la
> print 'Scipy version: ', __core_version__
> i = identity(4, Float64)
> print 'identity matrix:'
> print i
>
> print 'about to get cholesky decomposition'
> c = la.cholesky(i)
> print c
> ------------------------------------------------------------
>
> which gives
>
> ------------------------------------------------------------
> Scipy version: 0.6.1
> identity matrix:
> [[ 1. 0. 0. 0.]
> [ 0. 1. 0. 0.]
> [ 0. 0. 1. 0.]
> [ 0. 0. 0. 1.]]
> about to get cholesky decomposition
> Traceback (most recent call last):
> File "/Users/jaffe/Desktop/bad_cholesky.py", line 13,
>in ?
> c = la.cholesky(i)
> File
>"/Library/Frameworks/Python.framework/Versions/2.4/lib/
> python2.4/site-packages/scipy/linalg/basic_lite.py",
>line 117, in
> cholesky_decomposition
> results = lapack_routine('L', n, a, m, 0)
> lapack_lite.LapackError: Parameter a is not contiguous
>in
> lapack_lite.dpotrf
> ------------------------------------------------------------
>
> (The cholesky decomposition in this case should just be
>the matrix
> itself; the same error occurs with a complex matrix.)
>
>
> Any ideas? Could this have anything to do with
>_CastCopyAndtranspose
> in Basic_lite.py? (Since there are few other candidates
>for anything
> that actually changes the matrix.)
>
> Thanks in advance,
>
> A
>
>
> ______________________________________________________________________
> Andrew Jaffe
> a.jaffe at imperial.ac.uk
> Astrophysics Group
> +44 207 594-7526
> Blackett Laboratory, Room 1013 FAX
> 7541
> Imperial College, Prince Consort Road
> London SW7 2AZ ENGLAND
> http://astro.imperial.ac.uk/~jaffe
>
>
> _______________________________________________
> Scipy-dev mailing list
> Scipy-dev at scipy.net
> http://www.scipy.net/mailman/listinfo/scipy-dev
Scipy version: 0.7.1.1530
identity matrix:
[[ 1. 0. 0. 0.]
[ 0. 1. 0. 0.]
[ 0. 0. 1. 0.]
[ 0. 0. 0. 1.]]
about to get cholesky decomposition
[[ 1. 0. 0. 0.]
[ 0. 1. 0. 0.]
[ 0. 0. 1. 0.]
[ 0. 0. 0. 1.]]
Nils
More information about the SciPy-Dev
mailing list