Closed chhongs closed 4 years ago
The code produces multiple instances of the same class, sometimes indicating there is a bug sometimes not. Happens because the metrics are computed for each commit and each affected file in the commit.
Couldn't find a way to do this, yet. Perhaps extend the ML pipeline to preprocess the data accordingly.
Required functionalities have implemented in both java code level and machine learning preprocessing level.
The code produces multiple instances of the same class, sometimes indicating there is a bug sometimes not. Happens because the metrics are computed for each commit and each affected file in the commit.