Plot outcome distribution across a design variable

plot_design() visualizes the contingency table stored by the design audit for a categorical batch or covariate variable. Empty cells are outlined so confounding and complete separation are visible at a glance.

Usage

plot_design(audit, variable = NULL, type = c("count", "proportion"))

Arguments

audit: A moat_audit object.
variable: Optional single string naming a categorical variable audited by check_design(). When NULL, the first audited batch variable is used; if no batch variable is available, the first categorical covariate is used.
type: A single string. "count" plots raw sample counts and "proportion" plots proportions within each level of variable.

Value

A ggplot2::ggplot() object.

Examples

data("toy_moat")
audit <- moat(toy_moat, outcome = "outcome", batch = "batch", n_perm = 99)
plot_design(audit, variable = "batch")