[Scipy-tickets] [SciPy] #1304: lobpcg fails on some machines for trivial problems
SciPy Trac
scipy-tickets@scipy....
Wed Oct 13 15:29:55 CDT 2010
#1304: lobpcg fails on some machines for trivial problems
--------------------+-------------------------------------------------------
Reporter: nicki | Owner: somebody
Type: defect | Status: new
Priority: normal | Milestone: 0.9.0
Component: Other | Version: 0.8.0
Keywords: |
--------------------+-------------------------------------------------------
Hi,
the attached code fails on my machine with
==================== *snip* ====================
Solving generalized eigenvalue problem with preconditioning
matrix size 5
block size 1
No constraints
iteration 0
[ True]
current block size: 1
eigenvalue: [ 25.]
residual norms: [ 120.]
iteration 1
[ True]
current block size: 1
eigenvalue: [ 0.]
residual norms: [ 1.03680013e+09]
iteration 2
[ True]
current block size: 1
eigenvalue: [ 0.]
residual norms: [ 2.39610086e+45]
iteration 3
Warning: overflow encountered in multiply
[ True]
current block size: 1
eigenvalue: [ 0.]
residual norms: [ inf]
Traceback (most recent call last):
File "logpcg_test.py", line 8, in <module>
lobpcg( A, X, verbosityLevel = 10 )
File "/usr/lib64/python2.6/site-
packages/scipy/sparse/linalg/eigen/lobpcg/lobpcg.py", line 391, in lobpcg
aux = b_orthonormalize( B, activeBlockVectorR )
File "/usr/lib64/python2.6/site-
packages/scipy/sparse/linalg/eigen/lobpcg/lobpcg.py", line 130, in
b_orthonormalize
gramVBV = sla.cholesky( gramVBV )
File "/usr/lib64/python2.6/site-
packages/scipy/linalg/decomp_cholesky.py", line 66, in cholesky
c, lower = _cholesky(a, lower=lower, overwrite_a=overwrite_a,
clean=True)
File "/usr/lib64/python2.6/site-
packages/scipy/linalg/decomp_cholesky.py", line 16, in _cholesky
a1 = asarray_chkfinite(a)
File "/usr/lib64/python2.6/site-packages/numpy/lib/function_base.py",
line 547, in asarray_chkfinite
"array must not contain infs or NaNs")
ValueError: array must not contain infs or NaNs
==================== *snap* ====================
whereas on another one, one receives the expected output
==================== *snip* ====================
Solving generalized eigenvalue problem with preconditioning
matrix size 5
block size 1
No constraints
iteration 0
[False]
final eigenvalue: [ 1.]
final residual norms: [ 0.]
==================== *snap* ====================
Cheers,
Nico
--
Ticket URL: <http://projects.scipy.org/scipy/ticket/1304>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.
More information about the Scipy-tickets
mailing list