-
Notifications
You must be signed in to change notification settings - Fork 9
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
small speedboost by precalculating conjugation #165
Conversation
Minimum allowed coverage is Generated by 🐒 cobertura-action against 7b61153 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks reasonable to me, couple points:
- one surprise with performance
- this breaks API so should invoke version 0.7.0 instead of the current next release of 0.6.0
- One line was changed that was not covered in our current tests, please add a test to cover that line 😄
Co-authored-by: Sander Roet <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
still 1 missing test, LGTM otherwise
Woops this didn't fix it yet |
I set precision to 4 for the standard deviation. interpolation of the spherical mask upon rotation slightly changes the mask, meaning the eventual scores won't be identical. It should still be very close and that is indeed true. |
Its fixed now. The lines are hit! |
Needed to merge changes from main to make it mergeable. Lets see if they now pass. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, feel free to merge on green
PR to improve the speed a bit more and setup a few things to allow easier implementation of phase randomization.
Conjugated rft's of the tomogram are precalculated (instead of template conjugation). This requires some different shifting of the output score map and a different mask for calculating the score map standard deviation (hence some changes there).
Results are consistent with previous implementation and all the tests pass.
also closes #103