Plot Jaccard subgroup similarity index (SSI) for causal distillation tree objects

The Jaccard subgroup similiarity index (SSI) is a measure of the similarity between two candidate partitions of subgroups. To select an appropriate teacher model in CDT, the Jaccard SSI can be used to select the teacher model that recovers the most stable subgroups.

Usage

plot_jaccard(...)

Arguments

...

Two or more causal distillation tree objects, each is typically the output of causalDT. Arguments should be named (so that they are properly labeled in the resulting plot).

Value

A plot of the Jaccard SSI for each tree depth.

Examples

n <- 50
p <- 2
X <- matrix(rnorm(n * p), nrow = n, ncol = p)
Z <- rbinom(n, 1, 0.5)
Y <- 2 * Z * (X[, 1] > 0) + X[, 2] + rnorm(n, 0.1)

cdt1 <- causalDT(X, Y, Z)
cdt2 <- causalDT(X, Y, Z, teacher_model = rboost)
plot_jaccard(`Causal Forest` = cdt1, `Rboost` = cdt2)