Dear PYMC3 community, I am trying to study do a posterior_predictive test of a simple inference problem but the model has 37 dimensions (‘resid’) and I am not sure how to get the 37 posterior_predictive plots. This is my model: beta_prior = 4 sd_prior = 1 coords = { 'resid': df_3fb5.columns, …

This can be done using the flatten argument of plot_ppc. You should pass an empty list as flatten to keep each resid as a different subplot. There is one example in the API docs https://arviz-devs.github.io/arviz/generated/arviz.plot_ppc.html#arviz.plot_ppc on how is flatten used. Unfortunately how…

Plot posterior predictive several coordinates

Questions

sperezconesa September 23, 2020, 9:18am 7

Great! I look forward to it.
In any case, should I post arviz questions here or on the github? This is technically a pymc3 forum.

Topic		Replies	Views
Cannot plot_posterior with several dimensions Questions	1	385	January 8, 2021
PyMC3 posterior prediction example; how to incorporate a matrix of betas? Questions	24	2943	April 23, 2021
Plotting Sample_Posterior_Predictive output with Arviz v5 arviz	6	1857	January 26, 2023
How do I assign coords to do a ppc for specific groups in my data? v5 arviz	13	1221	July 17, 2023
Plot posterior predictive results scatterplot in specific format using arviz version agnostic arviz	3	547	July 11, 2023

Plot posterior predictive several coordinates

Related topics