Comprehensive Prior Diagnostics

Comprehensive Prior Diagnostics

Usage

DPprior_diagnostics(
  fit,
  thresholds = c(0.5, 0.9),
  warning_policy = NULL,
  M_verify = NULL,
  abs_tol = 1e-10,
  rel_tol = 1e-08,
  allow_approximate = FALSE
)

Arguments

fit

A canonical dpprior.result/1 fit. Legacy and flat lists are rejected; call upgrade_DPprior_object() explicitly first.

thresholds

Reserved fixed summary-view thresholds. The only accepted value is c(0.5, 0.9). Use prob_wsb_exceeds() for arbitrary W_SB tails or wmax_tail_bounds() for certified W_max bounds.

warning_policy

Optional ordinary list with exact fields estimand, direction, weight_threshold, and action_threshold. W_max policies are retained as indeterminate because canonical diagnostics make no W_max point-probability claim.

M_verify

Optional independent quadrature order.

abs_tol, rel_tol

Comparison tolerances.

allow_approximate

Return an approximate bundle when TRUE; otherwise signal a typed condition carrying the complete bundle.

Value

A DPprior_diagnostics object using the shared result schema.

See also

Other diagnostics: DPprior_error_bounds(), compute_error_landscape(), compute_weight_diagnostics()