-
Notifications
You must be signed in to change notification settings - Fork 3
While following the tutorial about triaxial DEM test, I encountered an error during the execution of the code. The issue I faced is "raise np.linalg.LinAlgError(msg) numpy.linalg.LinAlgError: When allow_singular is False, the input matrix must be symmetric positive definite". If anyone else has encountered similar problems, or if you have suggestions for how to resolve this error, please feel free to share your insights. I’m currently stuck and any guidance would be greatly appreciated.
All reactions
Replies: 1 comment
Hi, sorry for seeing this too late. Generally, this error is caused by a zero covariance matrix. Do you have a zero in your calibration.system.obs_data? I would suggest setting calibration.inference.Bayes_filter.scale_cov_with_max = True. Please note there has been a recent update in the repo. We have been working towards a new release.