Visual predictive check plot for an exposure-response model

Compares observed response rates against simulated response rates from a model, stratified by a grouping variable. This function is model-agnostic: it operates purely on data frames. The sim data frame is expected to contain one row per simulated observation per replicate, with a sim_id column identifying each replicate (see e.g. erglm::erglm_vpc_sim() for one way to generate such simulations from a fitted model).

Usage

er_vpc_plot(data, sim, exposure, response, group_by, conf_level = 0.95)

Arguments

data: Observed data
sim: Simulated data, with the same exposure/response/group_by columns as data, plus a sim_id column identifying each replicate
exposure: Exposure variable (one variable, unquoted)
response: Response variable (one variable, unquoted). Assumed to be binary (0/1)
group_by: Variable (unquoted) to stratify predictions
conf_level: Confidence level

Value

A ggplot2 object

Examples

if (FALSE) { # \dontrun{
library(erglm)
mod <- erglm_model(ae2 ~ aucss + sex, erglm_data, family = binomial())
sim <- erglm_vpc_sim(mod)
er_vpc_plot(erglm_data, sim, aucss, ae2, group_by = aucss)
er_vpc_plot(erglm_data, sim, aucss, ae2, group_by = sex)
} # }