[SciPy-dev] problem with linalg.cholesky?

Nils Wagner nwagner at mecha.uni-stuttgart.de
Mon Nov 28 14:25:08 CST 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