Builds an exposure-response plot for a fitted model

Usage

er_plot(data, exposure, response, stratify_by = NULL)

er_plot_style(object, labels)

er_plot_show_model(
  object,
  model,
  keep_strata = NULL,
  style = "ribbonline",
  conf_level = 0.95
)

er_plot_show_quantiles(
  object,
  keep_strata = NULL,
  style = "errorbar",
  bins = 4,
  conf_level = 0.95
)

er_plot_show_datastrip(
  object,
  keep_strata = NULL,
  style = "jitter",
  panel = "both"
)

er_plot_show_groups(
  object,
  group_by,
  style = "boxplot",
  bins = NULL,
  keep_strata = NULL
)

er_plot_build(object)

Arguments

data: Observed data
exposure: Exposure variable (one variable, unquoted)
response: Response variable (one variable, unquoted)
stratify_by: Stratification variable used for color and fill (one variable, unquoted)
object: Partially constructed plot (has S3 class er_plot)
labels: Named list of labels
model: A fitted exposure-response model. Must implement er_predict(); implementing er_simulate() and er_summary() enables additional visualisations (see er_model_interface)
keep_strata: Logical, indicating whether this component should keep the color stratification
style: Character string used to specify the partial builder for this component
conf_level: Confidence level
bins: Number of exposure bins (not counting placebo)
panel: Character string: "upper", "lower", or "both" (the default)
group_by: Grouping variables to define groups for distribution plots (a tidyselection of variables)

Value

Plot object of class er_plot

Examples

if (FALSE) { # \dontrun{
library(erglm)
mod <- erglm_model(ae1 ~ aucss, erglm_data, family = binomial())

erglm_data |>
  er_plot(aucss, ae1) |>
  er_plot_show_model(mod) |>
  er_plot_show_quantiles() |>
  er_plot_show_groups(aucss) |>
  plot()

mod2 <- erglm_model(ae2 ~ aucss + sex, erglm_data, family = binomial())
plt <- erglm_data |>
  er_plot(aucss, ae2, stratify_by = sex) |>
  er_plot_show_model(mod2, keep_strata = FALSE) |>
  er_plot_show_quantiles() |>
  er_plot_show_datastrip() |>
  er_plot_show_groups(group_by = c(aucss, treatment), keep_strata = FALSE)

print(plt)
plot(plt)
} # }