Skip to contents

Plot Pooled AUROC and AUPRC Boxplots Across Multiple Folders

Source: R/machine_learning.R
get_pooled_boxplots.Rd

This function aggregates AUROC and AUPRC metrics from multiple folders (typically corresponding to different cohorts or models), and generates boxplots comparing model performance across groups.

Usage

get_pooled_boxplots(folder_paths, file_name, width = 12, height = 8)

Arguments

folder_paths

Character vector. Paths to folders containing .rds files with ML model performance results.

file_name

Character. Prefix for the saved PDF files containing the plots.

width

Numeric. Width of the saved plots in inches. Default is 12.

height

Numeric. Height of the saved plots in inches. Default is 8.

Value

Saves two PDF files in the Results/ directory:

  • Boxplots_AUROC_performance_<file_name>.pdf

  • Boxplots_AUPRC_performance_<file_name>.pdf

No object is returned to the R environment.

Details

Each .rds file is expected to contain a list object with a result$AUC element, which includes numeric values for both AUROC and AUPRC. Folder names are used as grouping labels in the plots.

Red dashed lines are drawn at a fixed reference value (e.g., 0.7) for visual interpretation.