Added multivariate sampling to randvars.normal #858
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In a Nutshell
This pull request adds an alternative sampling method to randvars.normal which consists of sampling standard normal variables and applying bias and covariance shift manually.
Detailed Description
The method assumes the mean as a vector and the covariance as a two-dimensional object as well as having cov_cholesky available. It samples standard normal random variables in the size of the mean vector and multiplies them with cov_cholesky and adds the mean vector. This allows scaling with the memory/computational efficiency of the implemented Operator classes, which was not possible before since sampling called the operation to_dense on the covariance.
Furthermore, two tests are added that check the mean against a zero covariance and test the shape when using a BlockDiagonalOperator as the covariance.
Related Issues
No issues were opened.