Testability prediction dataset

This is a dataset used in the paper 'An ensemble meta-estimator to predict source code testability' published in the Applied Soft Computing journal.
This dataset contains all experimental data used in our machine learning pipeline at the ADAFEST project. The complete implementation is available on GitHub: https://m-zakeri.github.io/ADAFEST/.
Our testability prediction dataset consists of several CSV files which differ in preprocessing steps used to generate them. Each row denotes metrics for a Java class. Each column is a source code metric or test criteria obtained by running EvoSuite on the corresponding class under test. The first column is a long name (package_name.class_name) of a Java class.