PropensityScore.build#

PropensityScore.build(X, y, coords=None, prior=None, noncentred=True)[source]#

Construct the propensity graph without sampling.

Parameters:

  • X (ndarray) – Covariate matrix used to predict the treatment.

  • y (ndarray) – Observed treatment indicator (0/1); named t elsewhere.

  • coords (dict[str, Any] | None) – Coordinate names for named dimensions of the model.

  • prior (dict[str, Any] | None) – Historical compatibility parameter; currently unused by build_model(), which reads priors.

  • noncentred (bool) – Reserved for future non-centred parameterisations of the coefficient prior. Currently informational only.

Return type:

None