lmcinnes / lmcinnes/umap

UMAP segmentation fault

Open
#730 8 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
8.3k
Forks
871
Avg merge
1d 13h
Merged PRs (30d)
5

Description

```
umap.UMAP(n_neighbors=11, metric='jaccard',
min_dist=0.5).fit(X)
```
Error - segmentation fault

umap-learn 0.5.1 - Bad
umap-learn 0.5.0 - Bad
umap-learn 0.4.6 - Good

But if you increase it `n_neighbors` to 500, then everything works on new versions.

Contributor guide

Open the contributing guide

Research direction

Start by reproducing the provided umap.UMAP(...).fit(X) example with umap-learn 0.5.1, 0.5.0, and 0.4.6, comparing n_neighbors=11 with n_neighbors=500. Investigate the regression that causes the segmentation fault and verify that the failing case completes without crashing while preserving the working behavior.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.