This is Version 2 (early development) of the MVGC Multivariate Granger Causality MATLAB software suite.
The MVGC2 project will enhance and extend the existing open-source MVGC Multivariate Granger Causality MATLAB toolbox (see github.com/SacklerCentre/MVGC1). This project will update the current MVGC: (i) to take advantage of new state-of-the-art state-space and spectral methods developed in-house; (ii) to implement the MVGC functionality in Python (MVGC2-P); and (iii) to integrate the MVGC functionality with the popular neuroimaging software suites EEGLAB (MVGC2-EEGLAB) and MNE (MVGC2-MNE). The enhanced algorithms, as well as improving on efficiency and accuracy, also address some well-known problems associated with standard Granger causality inference from neuroimaging data. The Python implementation and EEGLAB/MNE integration facilitate users to deploy the MVGC functionality within their standard toolchains.
Currently, this repository contains the MVGC2 standalone MATLAB toolbox. To use the toolbox, first inspect the config.m script in the root directory, and edit to taste.
Make sure to run the startup.m script (e.g., by starting MATLAB in th MVGC2 root directory).
WARNING: Do NOT add the entire MVGC2 directory tree to your MATLAB path - this will cause things to break! MVGC2 paths are set appropriately by the startup script.