Replication package for "Do they work as reported? Reproducing and re-examining benchmarks for learning-based test case prioritization in continuous integration"

This package reproduces the numbers, tables and figures of the paper above. It contains the analysis scripts, per-cycle results and statistical tables, the training configurations and the patches applied to the original code, and scripts that fetch the third-party code and data from their original sources. Code is released under the MIT license; data and results under CC BY 4.0. README.md describes the layout and the command that regenerates each table and figure.