MMMSummaryFactory.prior_vs_posterior#

MMMSummaryFactory.prior_vs_posterior(var_names=None, dims=None, num_points=200, output_format=None)[source]#

Create prior vs posterior density grid summary DataFrame.

Evaluates Gaussian KDEs on a fixed grid for stable JSON export.

Parameters:
var_nameslist[str], str, or None, optional

Variables to summarize. None uses variables present in both prior and posterior groups.

dimsdict, optional

Coordinate filters applied before flattening samples.

num_pointsint, default 200

Number of evaluation points on the density grid.

output_format{“pandas”, “polars”}, optional

Output DataFrame format (default: uses factory default)

Returns:
pd.DataFrame or pl.DataFrame

Summary DataFrame with columns:

  • variable: Parameter name

  • x: Evaluation point on the density grid

  • density_prior: Prior KDE density at x

  • density_posterior: Posterior KDE density at x

  • <custom_dims>: Facet dimensions when present