andrija-djurovic/PDtoolkit ? reverse-engineered prompt

Build me an R package for credit risk modeling that helps with probability of default model development and validation.

I want it to take a loan data set and make it easy to do the usual modeling steps, like checking the data, handling missing values, binning numeric and categorical variables, calculating WoE and information value, coding variables with WoE, clustering risk factors, and measuring things like AUC. It should also include validation checks for model homogeneity, heterogeneity, discriminatory power, and predictive power.

Please include a sample data set and an example workflow using it so I can see how the package is meant to be used from start to finish. If you need to look up current R package docs or best practices while building it, feel free to do that. I’d like the package to feel polished and practical for real world credit scoring work, not just a toy example.