The batched version (shape=10) can be considerably faster as it produces vectorized logp and gradient code so it’s strongly suggested over separating the terms.
It’s not just a convenience
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Can you specify different parameters for a vector of priors? | 9 | 1233 | April 6, 2022 | |
| For loop vs shape parameter | 2 | 966 | May 4, 2021 | |
| Multiple Linear Regression | 2 | 693 | April 26, 2019 | |
| PyMC3 slows rapidly with increasing numbers of parameters | 5 | 11144 | December 3, 2018 | |
| Advice on dimensionality reduction? | 10 | 954 | March 27, 2019 |