Expanding the bayesplot vignette library

[florence-bockting]

Summary

While bayesplot currently offers a few introductory vignettes, the user experience could be improved by more targeted documentation. In the following three suggestions:

Suggested Vignettes

1. "Which plot should I use?"

This problem-oriented vignette would provide a structured overview (e.g., a lookup table) mapping specific diagnostic questions to their corresponding plot functions.

Diagnostic Question	Recommended Function(s)
Did my MCMC chains mix?	mcmc_trace(), mcmc_rank_overlay()
Is my R-hat acceptable?	mcmc_rhat(), mcmc_rhat_hist()
...	...

2. "Bayesplot and the Bayesian Workflow"

A workflow-oriented vignette that walks users through a complete modeling cycle using a single running example. It would demonstrate the appropriate plots for each stage.

3. "Customizing Bayesplot Outputs"

Since bayesplot returns standard ggplot objects, they can be modified using standard ggplot2 functions. However, it might be that some users are unaware of this extensibility. This vignette would guide users on how to benefit from this composability, covering topics such as: Adding layers or changing labels using the + operator, utilizing built-in color schemes and themes, customizing plot appearance via ggplot2 theme elements.

Relation to other issues

This issue is related to Issue#144 (Extend and Update vignettes) and Issue#142 (Vignette/template demonstrating how to contribute a new plotting function to bayesplot).

[jgabry]

I agree these would be good additions