Yellowbrick

Yellowbrick is an open source Python library for machine learning visualization. It extends the Scikit Learn API with a new core object called the Visualizer, which fits and transforms inside a normal Scikit Learn pipeline and draws its results with Matplotlib. The aim is to let data scientists see what a model is doing during selection and tuning, rather than relying only on automated grid search. It is built for data scientists and machine learning engineers who already work in Scikit Learn. The visualizers cover several stages of the workflow. Feature analysis includes Rank Features, Parallel Coordinates, Radial Visualization, PCA Projection and Manifold Visualization. Classification tools include a Classification Report, ROC and AUC curves, Precision Recall curves, Confusion Matrices and Discrimination Threshold. Regression tools include Prediction Error and Residuals plots, Alpha Selection and Cook's Distance. Output is described as publication ready, and the figures remain open to fine grained Matplotlib control. The documentation provides a Quick Start, a Model Selection Tutorial and one line quick methods. The project accepts contributions and follows the Python Software Foundation Code of Conduct. No commercial pricing is published, as the library is distributed as open source software.