spatial_autocorr crashes when the graph contains singletons
@timtreis is already working on this.
Since Aug 18, 2026.
- Dominant language
- Python
- Stars
- 598
- Forks
- 121
- Avg merge
- 3d 11h
- Merged PRs (30d)
- 3
Description
Report
Hi,
I thought I'd document something that i ran into. Unfortunately I don't have the logs anymore but let me know and I can try to reproduce if necessary.
I ran sq.gr.spatial_autocorr and it crashed with a seg fault. Upon debugging I realized that I had added an edge pruning step to my graph building, which led to some cells not having any neighbor anymore. If I'm not mistaken, there's a normalization step in spatial_autocorr, performed on the spatial weights matrix. Singletons get a sum of weights of 0 which cascades into a seg fault downstream.
A more transparent error upon detecting singletons would definitely help. Alternatively, the function could be filtering them out automatically?
I'm currently running this step on a filtered version of the AnnData, then exporting the scores separately so I haven't been bothered by not having the scores stored in .uns
Cheers!
Versions
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.
Assessment
This issue has not been assessed yet.