OpenMDAO may be doing unnecessary linear solver iterations when computing coloring.
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 782
- Forks
- 314
- Avg merge
- 4d 21h
- Merged PRs (30d)
- 4
Description
Description
In this StackOverflow issue: https://stackoverflow.com/questions/73666332/should-linear-solver-be-converging-when-coloring-is-beign-computed/73666657#73666657 the user was getting convergence failure warnings during coloring. It seems that we should be able to skip iterations in most cases during coloring since we're just using random numbers for our partials anyway. It might save us some time during the sparsity computation part of the coloring, which is the most expensive part of the process.
Example
No example.
OpenMDAO Version
3.19.1-dev
Relevant environment information
No response
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 reading the linked Stack Overflow discussion and tracing the coloring sparsity-computation path in OpenMDAO. Determine where linear solver iterations occur during coloring, then verify that unnecessary iterations are skipped without changing the computed coloring or introducing convergence warnings.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- performance
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100