Software

PyDaddy - Data Driven Dynamics in Python

PyDaddy is a Python package to automatically discover stochastic differential equation (SDE) models from time-series data. Given time-series data, PyDaddy discovers interpretable SDEs that best describe the underlying dynamics. The package also has a suite of visual diagnostic tools to let you diagnose the validity of the discovered models.

(The package was last updated in 2024 and is getting slightly old in the tooth. A version 2 with some performance and usability improvements is planned for some unspecified time in the future. The current version predates AI, so every line of code was meticulously handwritten by me or Ashwin.)

Bubblewrite - Write first, arrange later

Bubblewrite is a tool that I always wished I had while writing paper abstracts, project proposals etc. When I write, I like to put down individual points and ideas as separate sentences, and worry about how they arrange into a paragraph later. It’s a single HTML file that runs locally, entirely in your browser (except for optionally loading fonts from Google Fonts); no data leaves your computer. You can even save the html file onto your computer and run it entirely locally.

(Disclaimer: This was fully vibe-coded using Claude Code.)