Skip to content

Conversation

@bschneiderheinze
Copy link

Added classes for EkfdPredict and GmekfPredict to filters, and respective tests to tests. The tests should be comprehensive and they fully pass.

…h both pass. EkfdPredict test checked against manual ekf update, checked that a diagonal H implies a diagonal W, and checked against KfdPredict for linear measurements. GmekfPredict test checked against EkfdPredict for single mixand, checked that weights get heavily scaled down for non-agreeing measurements, and checked against GmkfPredict for linear measurements.
(ny,ny) measurement noise covariance matrix
H : ndarray or callable
(nz,nx) measurement Jacobian matrix
z_k = H(tk, xk, *args) @ x. If provided an ndarray instead, H will
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Here, we may want to force the user to specify H as a callable to avoid user errors caused by using an incorrect constant Jacobian

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I changed it to be consistent with the handling of F in EkfdPredict, where it is documented as a callable but if provided an ndarray is still recast as a callable. I can add an error/warning if we want that though.

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