You are not forced to match the size of the observed data in your posterior predictive samples. The default behaviour is to have 1 posterior predictive sample per 1 posterior sample but the documentation explains how the API is more flexible: pymc.sample_posterior_predictive — PyMC dev documentation
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Trouble understanding sample_ppc | 4 | 651 | June 7, 2018 | |
| Posterior Predictive Checks | 1 | 614 | May 19, 2019 | |
|
Posterior predictive checks with Gaussian Process
|
4 | 1330 | May 26, 2022 | |
| Sample_ppc shape | 5 | 2687 | July 18, 2018 | |
| Posterior predictive sampling with data variance | 10 | 2424 | September 14, 2018 |