summarise#

gpjax.summary.summarise(model, *, columns=None, console=None, max_array=4, precision=3, title=None, priors=None)[source]#

Print a rich table summarising a GPJax model’s parameters.

Renders one row per parameter – showing the constrained value, bijector, trainability, shape, and dtype – for any GPJax model (kernel, prior, posterior, likelihood, or variational family).

Parameters:

  • model (Any) – Any GPJax pytree (kernel, prior, posterior, likelihood, variational family, …).

  • columns (Sequence[str] | None) – Subset/ordering of columns to show. Defaults to the full GPflow-style column set.

  • console (Console | None) – Target rich.Console; defaults to a fresh one.

  • max_array (int) – Maximum number of array elements shown per value.

  • precision (int) – Significant figures for numeric values.

  • title (str | None) – Table title; defaults to the model’s class name.

  • priors (Mapping[str, Any] | None) – Optional mapping from a parameter’s name (as shown in the Parameter column) to a prior object (e.g. a NumPyro distribution), used to populate the Prior column. Defaults to None (all -).

Return type:

None

Example

>>> import gpjax as gpx
>>> kernel = gpx.kernels.RBF()
>>> gpx.summarise(kernel)

Expand for references to gpjax.summary.summarise

summarise