Posterior rank probabilities at target effect-modifier means

The full rank distribution behind cpaic_ranks(): the posterior probability that each treatment (or component) takes each rank, using average conditional link-scale effects evaluated at target means. Ported from multinma's posterior_rank_probs() (Phillippo et al. 2020) and extended because the component effects are beta + Gamma x, so the ranks move with x. These are not ranks of marginal standardized ORs, RRs, or HRs.

Usage

rank_probs(
  object,
  newdata = NULL,
  what = c("treatment", "component"),
  set = NULL,
  lower_is_better = FALSE,
  cumulative = FALSE,
  include_screen_only = FALSE,
  estimand = "average_conditional_link",
  ...
)

Arguments

object: A cmlnmr() fit.
newdata: A one-row data frame giving target effect-modifier means.
what: "treatment" (default) or "component".
set: Optional character vector restricting the elements to rank, as in cpaic_ranks(). Defaults to all treatments, including the reference, or all components.
lower_is_better: If TRUE, a smaller effect is preferred.
cumulative: Return cumulative rank probabilities (the quantity SUCRA summarizes) instead of the rankogram? Default FALSE.
include_screen_only: If FALSE (default), elements identified only by aggregate arms (a first-order screen) are excluded, as in cpaic_ranks().
estimand: The only implemented value is "average_conditional_link". Marginal standardized ranks are rejected.
...: Unused.

Value

A data frame of class cpaic_rank_probs with one row per (element, rank) and columns element, rank_position, and probability.

Details

Elements that are not estimable at the target means are dropped from the ranking set rather than ranked from the prior, exactly as in cpaic_ranks() (Step 3 of Wigle et al. 2026); they are listed in the dropped attribute.

References

Wigle A, Beliveau A, Nikolakopoulou A, Lin L (2026). Creating Treatment and Component Hierarchies in Component Network Meta-Analysis.

Examples

if (FALSE) {
rp <- rank_probs(fit, newdata = data.frame(x1 = 0.5), what = "component")
plot(rp)
}