Plot QINI Curves from Pre-computed Data — margot_plot_qini

Internal function to create QINI plots directly from pre-computed qini_data without requiring the full causal forest results structure.

Usage

margot_plot_qini_direct(
  qini_data,
  qini_objects = NULL,
  outcome_var = "Outcome",
  label_mapping = NULL,
  spend_levels = 0.1,
  show_spend_lines = TRUE,
  spend_line_color = "red",
  spend_line_alpha = 0.5,
  theme = "classic",
  show_ci = FALSE,
  ci_alpha = 0.05,
  ci_n_points = 20,
  ci_ribbon_alpha = 0.3,
  ci_ribbon_color = NULL,
  horizontal_line = TRUE,
  ylim = NULL,
  fixed_ylim = FALSE,
  cate_color = "#d8a739",
  ate_color = "#4d4d4d",
  treatment_cost = 1,
  x_axis = NULL,
  ...
)

Arguments

qini_data: Data frame with columns: proportion, gain, curve
qini_objects: List of maq objects (optional, for confidence intervals)
outcome_var: Character string for the outcome variable name
label_mapping: Optional named list for label transformations
spend_levels: Numeric vector of spend levels for vertical lines
show_spend_lines: Logical; show vertical spend lines
spend_line_color: Color for spend lines
spend_line_alpha: Alpha for spend lines
theme: ggplot2 theme name
show_ci: Logical or character for confidence intervals
ci_alpha: Significance level for CI
ci_n_points: Number of points for CI computation
ci_ribbon_alpha: Alpha for CI ribbons
ci_ribbon_color: Color for CI ribbons
horizontal_line: Logical; draw horizontal lines for complete paths
ylim: Y-axis limits
fixed_ylim: Logical; use fixed y-axis scaling
cate_color: Color for CATE curve
ate_color: Color for ATE curve
treatment_cost: Treatment cost value
x_axis: Type of x-axis: "proportion" or "budget". If not specified, will be inferred from data.
...: Additional arguments (ignored)

Value

A ggplot object