sweetviz

Why we included this project

Sweetviz is for the moment right before you fit a model, when you still need to understand what your data actually looks like. It reads a pandas DataFrame and produces a self-contained HTML report that covers missing values, distributions, and how the features correlate with one another. Its real strength is comparison: you can point it at a target column and see how every feature relates to it, or line up training and test sets side by side to catch leakage and drift before they cost you. The report is a single portable file, so it is easy to attach to a PR or pass to a teammate when you want evidence behind a data-quality call without standing up a dashboard. For routine tabular work in Python it is a quick, low-ceremony way to start an analysis.