CaTabRa: Efficient Analysis and Predictive Modeling of Tabular Data

We present CaTabRa, a novel open-source Python package for the efficient and largely automated analysis of tabular data. It combines a variety of established frameworks and libraries for data processing, automated machine learning, explainable AI and out-of-distribution detection into one coherent system. Thanks to its simple user interface, CaTabRa can be used by practitioners who want to quickly gain insights into their data and the potential of predictive modeling, but it also provides added value for data-science experts through its function library. We demonstrate CaTabRa’s usefulness in two example applications.