This project is a in-class study case to understand SINDy algorithm.
This sparse identificiation method is developed by Dr. Steven L. Brunton et.al
- citation1: Discovering governing equations from data by sparse identification of nonlinear dynamical systems, PNAS
- citation2: Inferring Biological Networks by Sparse Identification fo Nonlinear Dynamics, (Brunton Group, Niall Mangan), IEEE Transactions on Molecular, Biological and Multi-Scale Communications
Here, we transimitted this method from MATLAB to python, and also applied this method to several biological systems.
- Lorenz system
- Michaelis Menten (enzyme kinetics)
- SEIR (disease transmission model)
- RSM (antibiotic resistance gene transfer model)
Denoise algorithms and new information criteria of model selection were proposed in this code.
There are several documents is in good shape (final version):
- final_v1 folder, in which contains one utils.py and four .ipynb
- Utlis.py contains all functions/ algorithms that is utlized in system identification
- Four .ipynb, each one shows one example of the governing equation.
- figures folder, in which contains all figures we generate
- GBCB_5874_report.pdf, final report
Others are the reduancy files(drafts)