umap fitness metric
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 8.3k
- Forks
- 871
- Avg merge
- 1d 13h
- Merged PRs (30d)
- 5
Description
Is there an metric within UMAP which provides an overall scalar for the error in the embedding? Perhaps it could be useful to provide a function `umap_score(data, embedding)` in the API to call the loss function directly for use cases such as:
* Running UMAP on the same data generates slightly different results - is any better than the rest?
* If the data changes over time, can we measure the error introduced in keeping the previous embedding?
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by examining UMAP's existing loss function and API surface to determine whether an embedding score can be computed from the supplied data and embedding. Define what the scalar should measure for repeated runs and changing data, then verify that a proposed umap_score(data, embedding) interface answers both use cases consistently.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- data-visualization, machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100