lmcinnes / lmcinnes/pynndescent
Feature request: Other correlation metrics #319
- Dominant language
- Python
- Stars
- 970
- Forks
- 107
- PR merge metrics
- No merged PRs in 30d
Description
https://github.com/lmcinnes/umap/issues/319
@gokceneraslan
It'd be great to add other correlations as distances, such as Spearman's rho or the new correlation coefficient called phi_k. I am aware that it's possible to use any distance metric using either metric='precomputed' or implementing a custom function, but it'd be more convenient to pass metric='spearmanrho' or 'phik' directly.
Contributor guide
Research direction
Start by locating the code that handles the metric argument and existing distance metrics; inspect how metric='precomputed' and custom functions are supported. Find the tests covering metric selection, then define done as accepting the requested Spearman rho and phi_k metric names directly with matching distance behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 42/100