The observed part of a partially observed variable has no uncertainty or randomness whatsoever. It’s exactly the same as if you passed x as an array of numbers. The unobserved part is the one that will actually be sampled/inferred.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Making predictions on model inputs based on observed outputs | 3 | 751 | March 18, 2021 | |
| Sampling plausible posterior x parameter values in linear model for given y | 6 | 101 | July 30, 2024 | |
| The pymc3 way: Linear regression and inferring given posterior/trace | 4 | 1780 | May 9, 2018 | |
| Predicting from a model with multiple observed variables? | 4 | 4061 | October 9, 2018 | |
| Bayesian calibration on GP model | 7 | 1146 | July 23, 2022 |