Linear and nonlinear disambiguation optimization (LANDO)

This repository contains the codes used to perform the LANDO algorithm. The framework is described in the following schematic and reference

The code allows you to generate all the data used in the paper. Alternatively, you can download the data directly at this Dropbox link.

If you want to generate the data manually then you need to install Chebfun.

Here's a brief summary of the main codes. Further details can be found in the function files, or by calling, for example, "help trainLANDO".

File	Description
defineKernel.m	Defines a kernel cell based on provided hyperparameters
trainLANDO.m	Trains a LANDO model based on data, sparsification parameter and kernel
linopLANDO.m	Extracts and analyses the linear component of the LANDO model relative to a given base state
predictLANDO.m	Forms predictions based on a given LANDO model

To get started, add the source and example folders to the path (addpath('src','examples')) and try running lorenzExample.m

Please contact me if you encounter any bugs or have requests for features.

Reference:

Kernel Learning for Robust Dynamic Mode Decomposition: Linear and Nonlinear Disambiguation Optimization (LANDO)
Peter J. Baddoo, Benjamin Herrmann, Beverley J. McKeon & Steven L. Brunton
arXiv:2106.01510