mintyr - Grouped and Nested Data Pipelines Built on 'data.table'

A toolkit for grouped and nested data pipelines built on 'data.table': import many Excel / CSV files (including multi-row headers) into one table, reshape and nest data by trait and group, run reproducible (stratified) k-fold cross-validation inside every group, summarise groups with report-ready descriptive statistics, and write each piece back to its own file or sheet. Developed for animal breeding, where it prepares phenotypic files for 'ASReml-R', 'HIBLUP' or 'DMU', but useful for any multi-group, multi-variable analysis.