Week 3,4 GSOC 2024

This blog post covers my contributions to the project "Prior Elicitation (PreliZ)" with ArviZ during the third and fourth weeks of GSoC 2024.

Work Done

1. Write tests for the predictive_explorer function for PyMC and Bambi models.

2. Removed Sphinx Thebe functionality and introduced a tab layout to display examples.

3. Made changes to support pymc >= 5.16.0 and pytensor >= 2.23.0.

Add tests for predictive explorer

Tests using pytest have been added for both PyMC and Bambi models in the Predictive Explorer. To review the Predictive Explorer and its tests as a whole, you can refer to this Jupyter notebook.

Documentation Changes

Users encountered many issues with the live code run functionality using sphinx_thebe. We considered discarding it and replacing it with something more useful and easier to understand. As a result, we introduced a tab layout. Examples of this can be found here.

Support for pytensor >= 2.23.0

With the latest update to PyTensor, accessing the inputs of parent TensorVariables has changed. The inputs are now located at the leaf nodes of the graph, and accessing them requires using the ancestors function. This led to changes in methods such as get_model_information() and nc_parents(). I initially found this challenging, but the guidance from my mentors and the PyTensor documentation were very helpful.

To do in the coming week

With the mid-term evaluation and my college evaluation coming up, it’s going to be a hectic week for me.

Thank you for reading my post!