Interaction of parameter bounds and parameter priors
Nobody has claimed this yet.
- Dominant language
- No language data
- Stars
- 66
- Forks
- 16
- Avg merge
- 3h 2m
- Merged PRs (30d)
- 10
Description
There have been different interpretations of how parameter bounds and parameter priors interact. The current specifications don't address that explicitly.
My interpretation would be that if a prior distribution exceeds the bounds, the bounds truncate the prior and the probability density needs to be re-scaled accordingly.
The alternative would be, that the prior distribution is not affected by the parameter bounds. This seems weird, since if once really believes that those parameter values outside the bounds have a non-zero probability, then the bounds should have been chosen wider in the first place.
I'd be happy to get other's opinions on that matter.
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 reviewing the current specifications covering parameter bounds and parameter priors; no specific file, test, or entry point is identified in the issue. Compare the truncation and unaffected-prior interpretations, then clarify the specification so the interaction and any required probability-density rescaling are explicitly defined.
Written by the indexing model from the issue text.
Assessment
- Domain
- documentation
- Issue type
- Documentation
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100