Replace accidental O(N^2) BFRT with O(N log N). 13% better on NETLIB#96
Closed
chris-maes wants to merge 3 commits intoNVIDIA:branch-25.05from
Closed
Replace accidental O(N^2) BFRT with O(N log N). 13% better on NETLIB#96chris-maes wants to merge 3 commits intoNVIDIA:branch-25.05from
chris-maes wants to merge 3 commits intoNVIDIA:branch-25.05from
Conversation
chris-maes
added
non-breaking
Contributor
Author
|
This is still a work in progress. As I have not yet removed the old code, since I am still using it to run benchmarks on the MIPLIB test set. |
chris-maes
changed the title
4 tasks
Contributor
|
Should this draft PR be deleted since #192 exists? |
Contributor
Author
Yes. I will close. Thank you. |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The current bound-flipping ratio test (BFRT) makes an O(N) pass through the remaining breakpoints for each breakpoint resulting in O(N^2) run time, where N is the number of breakpoints.
This PR places the breakpoints in a heap in O(N) time. And then removes the minimum breakpoint O(1) and reorders the heap O(log N) for each of the N breakpoints. Thus, the run time should be O(N log N).
On the NETLIB LP test set this results in a 67% reduction of time spent in the BFRT and a 13% reduction in overall solve time.
The entering variable chosen by the new code should be the same as the previous code (except when ties are broken differently when constructing the heap).