Thanks to visit codestin.com
Credit goes to github.com

Skip to content

ENH Set Affinity propagation labels if any clusters were found - #22217

Merged
thomasjpfan merged 11 commits into
scikit-learn:mainfrom
Micky774:affinity_propagation_labels
Feb 2, 2022
Merged

ENH Set Affinity propagation labels if any clusters were found#22217
thomasjpfan merged 11 commits into
scikit-learn:mainfrom
Micky774:affinity_propagation_labels

Conversation

@Micky774

@Micky774 Micky774 commented Jan 15, 2022 •
edited
Loading

Copy link
Copy Markdown
Contributor

Reference Issues/PRs
Fixes #17657

What does this implement/fix? Explain your changes.
Current plan is to change behavior so that regardless of convergence, AffinityPropagation returns its generated cluster centers (if any) and labels. To do this I will change documentation (and warnings) to emphasize that un-converged results may be degenerate and should be accepted at the user's own risks.

Any other comments?
Open to suggestions on behavior changes

@Micky774
Initial attempt to modify sklearn/cluster/_affinity_propagation
c7b071b 
- Changed behavior to output any clusters/labels generated
even if not fully converged
- Modified documentation to reflect this behaviour
@Micky774 Micky774 changed the title [WIP] Affinity propagation labels ENH Affinity propagation labels Jan 15, 2022
thomasjpfan
thomasjpfan reviewed Jan 15, 2022
View reviewed changes

@thomasjpfan thomasjpfan left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the PR @Micky774 ! Implementation looks good. I left a comment about backward compatible for core-devs to decide on.

Comment thread sklearn/cluster/tests/test_affinity_propagation.py
@thomasjpfan

thomasjpfan commented Jan 18, 2022

Copy link
Copy Markdown
Member

For the time being, we can add a whats new so changelog test passes.

@Micky774

Micky774 commented Jan 18, 2022 •
edited
Loading

Copy link
Copy Markdown
Contributor Author

For the time being, we can add a whats new so changelog test passes.

Done. I added it under v1.1. What is the default version to add updates to? Or rather, how can I determine it.

@thomasjpfan

thomasjpfan commented Jan 19, 2022

Copy link
Copy Markdown
Member

Done. I added it under v1.1. What is the default version to add updates to? Or rather, how can I determine it.

It's almost always the next minor release, which is currently 1.1. The only entries that go into 1.0 after 1.0 is released are bug fixes.

jeremiedbb
jeremiedbb approved these changes Jan 28, 2022
View reviewed changes

@jeremiedbb jeremiedbb left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm ok with the backward incompatible change. It's what we agreed on in the issue.

LGTM. Thanks @Micky774 !

thomasjpfan
thomasjpfan reviewed Jan 29, 2022
View reviewed changes
Comment thread sklearn/cluster/tests/test_affinity_propagation.py Outdated
@Micky774

Micky774 commented Feb 1, 2022

Copy link
Copy Markdown
Contributor Author

@thomasjpfan Just wanted to see if there were any other revisions you'd like to see, or if this is ready to merge

thomasjpfan
thomasjpfan approved these changes Feb 1, 2022
View reviewed changes

@thomasjpfan thomasjpfan left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Minor suggestion for the whats new. otherwise LGTM!

Comment thread doc/whats_new/v1.1.rst Outdated
@thomasjpfan thomasjpfan changed the title ENH Affinity propagation labels ENH Set Affinity propagation labels if any clusters were found Feb 1, 2022
@Micky774 @thomasjpfan
Update doc/whats_new/v1.1.rst
37c6f69 
Co-authored-by: Thomas J. Fan <[email protected]>
@thomasjpfan
thomasjpfan merged commit 27dda7a into scikit-learn:main Feb 2, 2022
@Micky774
Micky774 deleted the affinity_propagation_labels branch February 2, 2022 21:29
glemaitre pushed a commit to glemaitre/scikit-learn that referenced this pull request Feb 9, 2022
@Micky774 @thomasjpfan @glemaitre
ENH Set Affinity propagation labels if any clusters were found (sciki…
2571f17 
…t-learn#22217)

Co-authored-by: Thomas J. Fan <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@thomasjpfan thomasjpfan thomasjpfan approved these changes
@jeremiedbb jeremiedbb jeremiedbb approved these changes

Assignees

No one assigned

Labels

module:cluster

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Affinity Propagation fails on the datasets that were clusterable in version 21.3

3 participants

@Micky774 @thomasjpfan @jeremiedbb