REP

REP (Reproducible Experiment Platform) is a Python library designed for machine learning experimentation, offering a consistent interface to several ML libraries and tools for classification, regression, and model evaluation. It provides wrappers for TMVA, XGBoost, Sklearn, Theanets, Pybrain, and supports meta-learning, grid search, parallel computing, custom data storage, reproducibility features, and plotting. REP is ideal for data scientists and researchers who need an extensible, reproducible platform for developing and benchmarking machine learning models.