Johannes Brachem

Add functionality for plotting divergent transitions

1

When diagnosing divergent transitions, it can be helpful to plot them for diagnosing the problem. Example 1: Divergent transitions in pair-plots (red points are divergences):  Example 2: Divergent transitions...

Add a little helper function for creating transition functions without a group

1

As mentioned in #25 In the `01-lin-reg.qmd`, we still need to explicitly include the `_value` name part, because there, we do not have a group: ```python def draw_sigma(prng_key, model_state): a_prior...

DRAFT ... to be elaborated. ```python class Group: ... def kernels(self) -> list[Kernel]: ... ```

Tutorials are not included in the documentation reliably

2

I just took a look at our documentation page and was surprised to see that the tutorials were missing. This apparently happened after merging https://github.com/liesel-devs/liesel/pull/57 The `docs` action ran successfully...

Ideas for the next level of the DistRegBuilder

3

What we want to do: - [ ] Move distregbuilder to its own module - [ ] Move some `liesel.model.legacy` functionality over to the new distregbuilder module - [ ]...

This issue serves as a reminder and a tracker for the discussion about convenience functionality for predictions in Liesel. Currently, we do not really offer much in that regard. I...

Closes #194 The changes are: 1. The function now includes a progress bar 2. The function now avoids including the model state into the compiled functions

See this comment: https://github.com/liesel-devs/liesel/pull/198#issuecomment-2248741496 The issue in the CI does not originate from scipy, but instead from incompatibility of tensorflow probability to numpy 2.0. At first, I could not reproduce...

E.g. adding the progress bar.

This interface can be placed in between two jitted blocks, where the tqdm progress bar is updated at the moment. One usecase would be remote logging to weights and biases...