PyDaddy (Python Data-Driven Dynamics)
PyDaddy is a comprehensive and easy to use python package to discover data-derived stochastic differential equations from time series data. PyDaddy takes the time series of state variable \(x\), scalar or 2-dimensional vector, as input and discovers an SDE of the form:
where \(\eta(t)\) is uncorrelated white noise. The function \(f\) is called the drift, and governs the deterministic part of the dynamics. \(g^2\) is called the diffusion and governs the stochastic part of the dynamics.
An example summary plot generated by PyDaddy, for a vector time series dataset.
PyDaddy also provides a range of functionality such as equation-learning for the drift and diffusion functions using sparse regresssion and a suite of diagnostic functions.
Schematic illustration of PyDaddy functionality.
Get Started
To take PyDaddy for a walk, see the tutorial notebooks. The notebooks can be executed online on Google Colab; no installation is necessary!
To install PyDaddy on your system, see installation instructions.
See the usage guide and advanced usage tips for detailed instructions and tips on how to use PyDaddy.
Citation
If you are using this package in your research, please cite the associated paper as follows:
Nabeel, A., Karichannavar, A., Palathingal, S., Jhawar, J., Brückner, D. B., Danny Raj, M., & Guttal, V., “Discovering stochastic dynamical equations from ecological time series data”, arXiv preprint arXiv:2205.02645, to appear in The American Naturalist.