Feature request: a table/plot of K-L divergences
Nobody has claimed this yet.
- Dominant language
- R
- Stars
- 114
- Forks
- 31
- PR merge metrics
- No merged PRs in 30d
Description
Using projpred for variable selection, it strikes me as odd that it is easy to compare different sized models (e.g., via plot.vsel), but no way of comparing different variable sets for a given model size. For example, for a reference model of y ~ A + B + C, we can visualize the CV results (e.g., ELPD, RMSE) of models, y ~ A, y ~ A + B, and y ~ A + B + C, but we cannot see how models y ~ A + B, y ~ A + C, and y ~ B + C stack up against each other. It would be nice to be able to extract a table or plot of each model's K-L divergence values.
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 existing plot.vsel behavior and how projpred exposes cross-validation results and K-L divergence values. Determine how models with the same size but different variable sets should be represented, then define whether the completed feature returns a table, a plot, or both for comparing those values.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- r
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 38/100