Pycm

PyCM is a multi class confusion matrix library written in Python. It is used to evaluate classification models after training, accepting either input data vectors or a ready made matrix and then computing a broad set of per class and overall statistics from it. The project describes itself as the swiss army knife of confusion matrices. It is aimed at data scientists and machine learning engineers who need more metrics than a basic accuracy figure when judging a wide variety of classifiers, including cases with many classes. The library is installed with pip, and the site offers source downloads as a zip file or a tar.gz archive, a link to online documentation and a page for trying PyCM online in the browser. The current version shown on the home page is 4.6. PyCM was published in the Journal of Open Source Software in 2018, and the site asks that research using it cite that paper, with a BibTeX file provided. Development is credited to the PyCM Development Team. The library is distributed as free downloadable code, and the site lists no pricing, paid plans or hosted service.