numpy.linalg.linalg.LinAlgError: SVD did not converge during rmgpy.reaction.Reaction.reverse_arrhenius_rate #1833
Description
Bug Description
During calls to rmgpy.reaction.Reaction.reverse_arrhenius_rate I am now quite often getting numpy.linalg.linalg.LinAlgError: SVD did not converge in Linear Least Squares errors, that cause RMG to immediately crash.
One context was @ajuluc and I working on a tool to check collision limit violations.
Another was our chemkin mechanism importer script.
It might be that there are NAN values creeping in, but I don't see why that would happen in these contexts, and for the importer script it never used to be a problem, and it wasn't a problem when @ChrisBNEU ran the same input on his ubuntu machine on Wednesday, IIRC. So I suspect it might be an OS or compiler or hardware specific problem.
Based on some discussion at numpy/numpy#12941 etc.
Partly just sharing this here as a place for others to comment if they run into similar issues.
Here's some stack trace:
File "rmgpy/reaction.py", line 791, in rmgpy.reaction.Reaction.generate_reverse_rate_coefficient
File "rmgpy/reaction.py", line 872, in rmgpy.reaction.Reaction.generate_reverse_rate_coefficient
File "rmgpy/reaction.py", line 854, in rmgpy.reaction.Reaction.generate_reverse_rate_coefficient
File "rmgpy/reaction.py", line 829, in rmgpy.reaction.Reaction.generate_reverse_rate_coefficient
File "rmgpy/reaction.py", line 788, in rmgpy.reaction.Reaction.reverse_arrhenius_rate
File "rmgpy/kinetics/arrhenius.pyx", line 173, in rmgpy.kinetics.arrhenius.Arrhenius.fit_to_data
File "/Users/rwest/anaconda/envs/rmg3/lib/python3.7/site-packages/numpy/linalg/linalg.py", line 2156, in lstsq
x, resids, rank, s = gufunc(a, b, rcond, signature=signature, extobj=extobj)
File "/Users/rwest/anaconda/envs/rmg3/lib/python3.7/site-packages/numpy/linalg/linalg.py", line 101, in _raise_linalgerror_lstsq
raise LinAlgError("SVD did not converge in Linear Least Squares")
numpy.linalg.linalg.LinAlgError: SVD did not converge in Linear Least Squares
And here's my numpy config
$ python
Python 3.7.4 (default, Aug 13 2019, 15:17:50)
[Clang 4.0.1 (tags/RELEASE_401/final)] :: Anaconda, Inc. on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy as np
>>> np.show_config()
mkl_info:
libraries = ['mkl_rt', 'pthread']
library_dirs = ['/Users/rwest/anaconda/envs/rmg3/lib']
define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
include_dirs = ['/Users/rwest/anaconda/envs/rmg3/include']
blas_mkl_info:
libraries = ['mkl_rt', 'pthread']
library_dirs = ['/Users/rwest/anaconda/envs/rmg3/lib']
define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
include_dirs = ['/Users/rwest/anaconda/envs/rmg3/include']
blas_opt_info:
libraries = ['mkl_rt', 'pthread']
library_dirs = ['/Users/rwest/anaconda/envs/rmg3/lib']
define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
include_dirs = ['/Users/rwest/anaconda/envs/rmg3/include']
lapack_mkl_info:
libraries = ['mkl_rt', 'pthread']
library_dirs = ['/Users/rwest/anaconda/envs/rmg3/lib']
define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
include_dirs = ['/Users/rwest/anaconda/envs/rmg3/include']
lapack_opt_info:
libraries = ['mkl_rt', 'pthread']
library_dirs = ['/Users/rwest/anaconda/envs/rmg3/lib']
define_macros = [('SCIPY_MKL_H', None), ('HAVE_CBLAS', None)]
include_dirs = ['/Users/rwest/anaconda/envs/rmg3/include']
Installation Information
Describe your installation method and system information.
- OS: Mac OS 10.14.6
- Installation method: from source in Anaconda environment
- RMG version information:
- RMG-Py: 2.4.1-866-g66b751b4c (on importer branch)