Add BLINK score implementation #829
Merged
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
julianpollmann
approved these changes
Sep 12, 2025
julianpollmann
approved these changes
Sep 17, 2025
Collaborator
Author
|
Thanks a lot for reviewing @julianpollmann ! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
This is to have fast cosine estimates in matchms and is based on Harwood et al. (2023), but in an implementation that differs from the original GitHub implementation.
In terms of runtime both implementations are close. The original seems to still be a bit faster, but I considered that to not be worth spending too much time on (at least not to me right now).
One consequence of the different implementation is that the rounding is not the same for the tolerance windows. The original BLINK, say with a tolerance of
0.01was an approximation of theCosineGreedywith tolerance < 0.01 (which is why the authors usedtolerance=0.0099. The implementation in this PR, however, sticks to the same tolerance so it would approximate CosineGreedy with atolerance=0.01.