figure wishlist

[anne-urai]

Generate one generic script that plots useful diagnostics after each model fit

Create a new function (like this one) in the utils_plot.py script in the repo. Then each plot, indicating the model-dataset-figuretype is saved to a GitHub folder: here.

Diagnostics

• [ ] posterior predictives
• [x] plot_trace
• [ ] q-q plots
• [x] forest plot
• [x] marginal posteriors
• [x] pair plots

Then also figures that specifically replicate the eLife findings

• [ ] History-dependent chronometric and psychometric functions (already in https://github.com/anne-urai/mouse_history_ddm/tree/main/figures)
• [ ] Model comparison, see panel a. Code here
• [x] Correlation with P(repeat), see panel a. Code here

[anne-urai]

https://github.com/kiante-fernandez/2023_choicehistory_HSSM/blob/main/src/plot_figure4_posterior_choice_history.py