Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix Lorenz 63 model filtering bug #252

Conversation

matt-graham
Copy link
Member

Something seems to have changed in recent FillArrays / Distributions versions in the behaviour when setting mean and (co)variance parameters for MvNormal using FillArrays lazy array types Ones and Zeros rather than explicit array initializers ones and zeros, which appears to be leading to unintended interactions between particle updates when filtering. This PR switches to using the explicit array initializers and also initializes integrator state specifically to zero rather than undef to avoid numerical warnings from OrdinaryDiffEq.

@codecov-commenter
Copy link

codecov-commenter commented Dec 1, 2023

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (ee916ab) 91.42% compared to head (c12942d) 91.42%.

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files
@@           Coverage Diff           @@
##           master     #252   +/-   ##
=======================================
  Coverage   91.42%   91.42%           
=======================================
  Files           7        7           
  Lines         385      385           
=======================================
  Hits          352      352           
  Misses         33       33           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@matt-graham
Copy link
Member Author

Closing as should now be fixed in #256

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants