ENH Avoid repeated input checks in kmeans++ #19002
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
|
Future improvement could be to merge the computation of the euclidean distances with the computation of the min in a single function parallelized over data chunks in cython (a bit like it's done for the kmeans iteration loop. In fact maybe some part of the code could be shared). |
ogrisel
approved these changes
Dec 16, 2020
|
Please document it in whats new (for 1.0 as suppose as the previous perf was not catastrophic enough to be considered a bug). Although, since 0.24.0 is not yet out we could also consider it for inclusion in 0.24.0, as you wish. |
|
For reference, I made a summary of our current understanding of the perf issues with k-means++ init in #10924 (comment). |
thomasjpfan
approved these changes
Dec 17, 2020
Co-authored-by: Thomas J. Fan <thomasjpfan@gmail.com>
8000
Co-authored-by: Thomas J. Fan <thomasjpfan@gmail.com>
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Outdated suggestions cannot be applied.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.
kmeans++ calls
euclidean_distancesmany times and each call repeat the check of the input which is not necessary since it has already been checked.partially take over #7383
The time wasted in the input check can be huge on machines with many cores. For instance I tested on a machine with 44 cores. Here's the time spent during KMeans.fit (in seconds):
cc @ogrisel