[SciPy-dev] problem with linalg.cholesky?
Andrew Jaffe
a.h.jaffe at gmail.com
Mon Nov 28 14:37:53 EST 2005
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
More information about the SciPy-Dev
mailing list