/EquationLearning.jl

Equation learning with uncertainty quantification: Code for VandenHeuvel et al. (2022) "Computationally efficient mechanism discovery for cell invasion with uncertainty quantification"

Primary LanguageJuliaMIT LicenseMIT

Equation Learning

This package contains code to perform equation learning with bootstrapping using Gaussian processes, as described in our paper https://doi.org/10.1371/journal.pcbi.1010599. Currently the method is only implemented for partial differential equations of the form

equation

but the ideas could easily be extended to much more complicated problems or other classes of problems. Please see the documentation for more details and instructions for use. In this documentation we also provide details for reproducing the results in our paper.

Installation

To install the package, you can use:

] add https://github.com/DanielVandH/EquationLearning.jl.git

in the Julia REPL. Note that the ] prefix is to enter the Pkg REPL.

Issues

Any suggestions, questions, and/or issues with the package should be given as an issue, or as an email to Daniel VandenHeuvel (this link opens an email to vandenh2@qut.edu.au). Issues are preferred.

At the time of writing, the Issues tab is essentially being used by myself as a to-do list. Feel free to comment or ask about these features as well.