Rath

RATH is an open-source data analysis and visualization tool that tries to automate most of the work involved in exploring a dataset. Instead of manually building charts or writing code, you upload your data and the tool does the initial heavy lifting: finding patterns, spotting unusual values, suggesting charts, and identifying which variables seem to influence each other.

There are two main modes. AutoPilot runs a fully automated pass over your data and presents findings with visualizations generated on its own, requiring just one click to get started. The Copilot mode is more interactive: you guide the exploration and RATH generates relevant recommendations based on what you appear to be looking for. There is also a natural language interface so you can type a question about your data and get a chart or answer back, similar to asking a colleague who already knows the dataset.

The tool includes a Data Wrangler for cleaning and transforming raw data before analysis. It can automatically suggest transformations, such as extracting patterns from text columns or handling missing values. Once your data is prepared, the AutoVis feature picks the chart type it judges will communicate the data most clearly, so you can focus on what the data says rather than on chart configuration.

A feature called Data Painter lets you explore data by literally coloring points on a chart with your mouse and then asking the tool to explain what those colored points have in common. There is also a causal analysis module that tries to map out cause-and-effect relationships between variables rather than just showing correlations, which can help when making decisions based on data.

RATH runs in the browser and has a hosted sandbox you can try without installing anything. The project is written in TypeScript, published under the AGPL license, and includes a standalone chart component called Graphic Walker that developers can embed in their own applications.