Rankogram and cumulative rank plot

Ported from multinma's plot.nma_rank_probs() (Phillippo et al. 2020). The rankogram gives the posterior probability of each rank; the cumulative version gives the probability of being ranked among the best k, whose normalized area is SUCRA.

Usage

# S3 method for class 'cpaic_rank_probs'
plot(x, y, ...)

Arguments

x

A cpaic_rank_probs object from rank_probs().

y

Unused, for compatibility with the plot() generic.

...

Unused.

Value

A ggplot object.

Details

Both are computed at named target effect-modifier means. They summarize the average conditional link-scale hierarchy, not a marginal hierarchy.

See also

rank_probs(), plot_rank_curve()

Examples

if (FALSE) {
plot(rank_probs(fit, newdata = data.frame(x1 = 0), what = "component"))
}