Build a per-node summary table across all nodes

For each node, the node's mechanism contributes summary rows - for a Gaussian process, the direct local effect of each parent (the fitted function's gradient at the parents' means, comparable to a regression coefficient) and the residual noise sd.

Usage

build_summary_table(model)

Arguments

model

A fitted KaguModel.

Value

A tibble with columns node, term, mean, sd, hdi_lower, hdi_upper.