investigate online counts of post-warmup divergences (or any column) during sampling
Open
Nobody has claimed this yet.
enhancement
i/o
interface
user experience
- Dominant language
- C++
- Stars
- 238
- Forks
- 106
- Avg merge
- 2h 29m
- Merged PRs (30d)
- 6
Description
Summary:
When CmdStan completes a run, it should report number of post-warmup divergences and number of iterations where maxtreedepth was exceeded, and it should track this effeciently, i.e., minimal memory overhead.
We want this for cmdstanpy/cmdstanr and lightweight wrappers to avoid reading in the entire outputs.
Description:
tee sample_writer in command.hpp:
Current Version:
v2.22.1
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 command.hpp and the sample_writer mentioned in the issue to understand where sampling output is written. Done means CmdStan reports post-warmup divergences and iterations exceeding max treedepth while tracking them with minimal memory overhead, without requiring wrappers to read all outputs.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- cli
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100