Skip to content

K means binning#51

Merged
abaranov25 merged 12 commits into
sintel-dev:masterfrom
abaranov25:K-Means-Binning
Apr 28, 2026
Merged

K means binning#51
abaranov25 merged 12 commits into
sintel-dev:masterfrom
abaranov25:K-Means-Binning

Conversation

@abaranov25
Copy link
Copy Markdown
Collaborator

@abaranov25 abaranov25 commented Dec 10, 2025 •
edited
Loading

Resolve #56
Added K-Means binning as a way to pre-process floats into scalars.

@abaranov25 abaranov25 marked this pull request as draft February 9, 2026 05:34
@abaranov25 abaranov25 marked this pull request as ready for review February 11, 2026 21:46
@abaranov25 abaranov25 mentioned this pull request Feb 11, 2026
Closed
@abaranov25 abaranov25 requested a review from sarahmish February 17, 2026 23:05
@abaranov25 abaranov25 self-assigned this Feb 19, 2026
sarahmish
sarahmish reviewed Mar 12, 2026
View reviewed changes
Copy link
Copy Markdown
Contributor

@sarahmish sarahmish left a comment

Choose a reason for hiding this comment

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

According to the issue you initially were calling it Cluster2Float and Float2Cluster. Is there a reason you changed it to Cluster2Scalar and Scalar2Cluster?

Comment thread sigllm/pipelines/detector/mistral_detector_kmeans.json Outdated
Comment on lines +10 to +11
"sigllm.primitives.transformation.Cluster2Scalar",
"sigllm.primitives.transformation.Cluster2Scalar",
Copy link
Copy Markdown
Contributor

@sarahmish sarahmish Mar 12, 2026

Choose a reason for hiding this comment

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

What is the purpose of the second Clustr2Scalar? Note that the jupyter notebook in the PR seems to use an older version of the pipeline with only one Clustr2Scalar.

Copy link
Copy Markdown
Collaborator Author

@abaranov25 abaranov25 Apr 21, 2026

Choose a reason for hiding this comment

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

@sarahmish good catch. I probably confused myself with scalars and floats, but I changed it back to original wording.

I believe there was an update in sigllm that requires y and y_hat to both be passed through scalar2float separately for the pipeline to work. Otherwise one is scaled by decimal and the other is not.

@abaranov25 abaranov25 merged commit 357a6a5 into sintel-dev:master Apr 28, 2026
11 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@sarahmish sarahmish sarahmish approved these changes

Assignees

@abaranov25 abaranov25

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Add a New K-Means Binning Primitive

2 participants

@abaranov25 @sarahmish