Skip to content

Unnecessary Transpose in predict/update? #30

@dsandber

Description

@dsandber

I noticed these two lines in ekf_update and ekf_predict:

_transpose(H, Ht, EKF_M, EKF_N);
_transpose(F, Ft, EKF_N, EKF_N);

Aren't H and F constants? Or is it reasonable for them to change during run-time? If they're constant, then the transpose of those two matrixes can be saved at initialization time. Perhaps a #define would make sense for those that think the memory sacrifice is worth the speed-up?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions