Plot QINI Curves Across Treatment Cost Scenarios
Source:R/margot_plot_qini_cost_sensitivity.R
margot_plot_qini_cost_sensitivity.Rd
Visualizes how QINI curves change with different treatment costs, showing the impact of budget constraints on optimal treatment allocation.
Usage
margot_plot_qini_cost_sensitivity(
cost_sensitivity_result,
model_name,
plot_type = c("overlay", "facet"),
show_baseline = TRUE,
colors = NULL,
title = NULL,
subtitle = NULL,
...
)
Arguments
- cost_sensitivity_result
Output from margot_qini_cost_sensitivity()
- model_name
Character string specifying which model to plot
- plot_type
Character; either "overlay" (all costs on one plot) or "facet" (separate panel for each cost). Default is "overlay".
- show_baseline
Logical; whether to show baseline curves. Default is TRUE.
- colors
Optional vector of colors for different cost scenarios. If NULL, uses a gradient from blue (low cost) to red (high cost).
- title
Optional plot title. If NULL, auto-generated based on model name.
- subtitle
Optional plot subtitle. If NULL, describes the cost scenarios.
- ...
Additional arguments passed to ggplot2 functions
Details
This function creates visualizations showing how QINI curves change with treatment cost. Lower costs result in steeper curves (more people can be treated cost-effectively), while higher costs result in shallower curves (only highest-effect individuals justify treatment).
The "overlay" plot type shows all cost scenarios on one plot with different colors, making it easy to compare curve shapes. The "facet" plot type creates separate panels for each cost, useful when curves overlap significantly.
Examples
if (FALSE) { # \dontrun{
# Run cost sensitivity analysis
cost_sens <- margot_qini_cost_sensitivity(
causal_forest_results,
costs = c(0.2, 0.5, 1, 2, 5)
)
# Overlay plot (default)
margot_plot_qini_cost_sensitivity(cost_sens, "model_anxiety")
# Faceted plot
margot_plot_qini_cost_sensitivity(
cost_sens,
"model_anxiety",
plot_type = "facet"
)
# Custom styling
margot_plot_qini_cost_sensitivity(
cost_sens,
"model_anxiety",
colors = c("darkgreen", "gold", "orange", "red", "darkred"),
title = "Treatment Cost Impact on Anxiety Intervention",
subtitle = "Lower costs enable treating more patients"
)
} # }