Skip to content

Issue with kmean_features #20

@liuzk233

Description

@liuzk233

First of all, thank you very much for sharing your excellent work and codebase. While using your kmean_features clustering function, I encountered an issue I’d like to kindly ask about:

Even when the input features remain identical and a fixed random_state is set, the number of unique clusters in the KMeans output labels is sometimes less than the specified n_clusters=100 (e.g., 95 or 98). This suggests that not all 100 clusters are actually populated.

My questions are:

1.Did you observe this same behavior during the development of the project?
2.Do you have any recommended approaches to ensure that all 100 clusters are consistently utilized?

Thank you again for your valuable contribution to the community. I greatly appreciate your time and guidance.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions