Software Release: CausalSynth Web Application for Interactive Causal Graph-Driven Sythetic Data Generation
CausalSynth is an interactive web-based tool for specifying causal graphs that in turn generate sythetic data with known causal relations.
We are excited to announce the launch of CausalSynth, an interactive web application designed to facilitate the generation of synthetic datasets based on user-defined causal relationships. CausalSynth offers a user-friendly graphical interface that allows users to define acyclic causal graphs, specify variable interdependencies, and produce datasets that reflect these causal interactions. The tool also includes data visualization capabilities that allow users to inspect and validate the generated data through various visualizations like scatter plots, bubble plots, histograms, and causal graphs.
CausalSynth is straightforward to set up and use. Hosted at CausalSynth, the application can be accessed directly via the web. For those interested in running CausalSynth locally, detailed installation instructions are available on the tool’s GitHub repository which can also be found linked to our Software page. Simply clone the repository, install the necessary dependencies, and start the development server to begin.
More information can also be found in our IEEE VIS 2024 poster “CausalSynth: An Interactive Web Application for Synthetic Dataset Generation and Visualization with User-Defined Causal Relationships” which is available in our publication library.
This first release of the CausalSynth web application was created by lead developer Zhehao Wang (in collaboration with Arran Zeyu Wang, David Borland, and David Gotz) and was made possible in part by the National Science Foundation under Award #2211845.
Icon designed by grafixpoint and downloaded from FlatIcon.com.