TBFPC.__init__#
- TBFPC.__init__(target, *, target_edge_rule='any', bf_thresh=1.0, forbidden_edges=None, required_edges=None)[source]#
Create a new TBFPC causal discovery model.
- Parameters:
- target
Variable name for the model outcome; must be present in the data used during fitting.
- target_edge_rule
Rule that controls which driver → target edges are retained. Options are
"any","conservative", and"fullS".- bf_thresh
Positive Bayes factor threshold applied during conditional independence tests.
- forbidden_edges
Optional sequence of node pairs that must not be connected in the learned graph.
- required_edges
Optional sequence of directed
(u, v)pairs that must be present in the learned graph asu -> v.