TuringLang / TuringLang/DynamicPPL.jl

Warn on `±Inf` in `check_model`

Open
#889 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement
Dominant language
Julia
Stars
286
Forks
41
Avg merge
1d 11h
Merged PRs (30d)
34

Description

https://github.com/TuringLang/DynamicPPL.jl/pull/888 modifies the check_model function such that it errors if any NaN's are encountered on the left-hand side of a tilde statement.

This issue is to record the idea that there could be some kind of warning / info message for ±Inf as well. However, this would depend on whether it's reasonable to have Inf as a value; if Inf is actually in the support of the distribution on the right, then it would be fine.

Anyone implementing this would have to be careful about doing it correctly for multivariate distributions, especially product distributions.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the check_model function and the changes in pull request #888. Determine how ±Inf values should be handled when Inf is in the right-hand distribution's support, including multivariate and product distributions. Done means check_model provides an appropriate warning or info message without incorrectly flagging valid values.

Written by the indexing model from the issue text.

Assessment

Tech stack
julia
Domain
machine-learning
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.