Composite Scoring via Principal Component Analysis of Ridit Scores

Implements 'PRIDIT' (Principal Component Analysis applied to 'RIDITs'), an unsupervised, nonparametric method for aggregating ordinal, categorical, and continuous indicators into a single interpretable composite score. Originally proposed by Brockett et al. (2002) for insurance fraud detection and extended to hospital quality measurement by Lieberthal (2008) and Lieberthal and Comer (2013) . The package provides: (1) low-level functions ridit(), PRIDITweight(), and PRIDITscore(); (2) a unified pridit() entry point returning a classed object with print, summary, 'autoplot', and 'coef' methods; (3) pridit_boot() for bootstrap confidence intervals on scores and weights; (4) a step_pridit() recipe step for out-of-sample scoring within the 'tidymodels' framework; and (5) pridit_longitudinal() for panel data, computing cross-period stability of scores and weights.

pridit

An R package that implements the PRIDIT (Principal Component Analysis applied to RIDITs) analysis system as described in Brockett et al. (2002).

Installation

From CRAN (recommended):

install.packages("pridit")

From GitHub (development version):

# Install devtools if you haven't already
install.packages("devtools")

# Install the PRIDIT package
devtools::install_github("rlieberthal/PRIDIT")

Description

This package provides three main functions for calculating and analyzing Ridit scores and PRIDIT scores:

• ridit() - Calculates Ridit scores for a given dataset using the method developed by Bross (1958) and modified by Brockett et al. (2002)
• PRIDITweight() - Applies Principal Component Analysis (PCA) to Ridit scores to calculate PRIDIT weights for each variable
• PRIDITscore() - Calculates final PRIDIT scores using the weights and ridit scores

Quick Start

library(pridit)

# Load your data (first column should be IDs)
data <- data.frame(
  ID = c("A", "B", "C", "D", "E"),
  var1 = c(0.9, 0.85, 0.89, 1.0, 0.89),
  var2 = c(0.99, 0.92, 0.90, 1.0, 0.93),
  var3 = c(1.0, 0.99, 0.98, 1.0, 0.99)
)

# Step 1: Calculate ridit scores
ridit_scores <- ridit(data)

# Step 2: Calculate PRIDIT weights
weights <- PRIDITweight(ridit_scores)

# Step 3: Calculate final PRIDIT scores
final_scores <- PRIDITscore(ridit_scores, data$ID, weights)

print(final_scores)

Data Format

Your input data should be structured as:

• First column: Unique identifiers (IDs)
• Remaining columns: Numerical variables to be analyzed
• All variables should be numeric (convert factors/categories to numeric values like 0,1 or 1,2,3,4,5)

Output

The final PRIDIT scores range from -1 to 1, where:

• The sign indicates class identity
• The magnitude indicates the intensity of that identity

References

• Bross, I. D. (1958). How to use ridit analysis. Biometrics, 14(1), 18-38. doi:10.2307/2527727
• Brockett, P. L., Derrig, R. A., Golden, L. L., Levine, A., & Alpert, M. (2002). Fraud classification using principal component analysis of RIDITs. Journal of Risk and Insurance, 69(3), 341-371. doi:10.1111/1539-6975.00018
• Lieberthal, R. D. (2008). Hospital quality: A PRIDIT approach. Health services research, 43(3), 988-1005.

License

This project is licensed under the Apache License 2.0.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.