Skip to content

CompBioUIC/DTWGrangerFramework

Repository files navigation

DTWGrangerFramework

The framework of VL-Granger causality inference

The main file of this framework is "DTW_granger_cause.m".

Varible-Lag Granger Causality using DTW

Developer: Chainarong Amornbunchornvej, UIC, 2018

Revised code based on [*] by Chandler Lutz We use the HSIC Independence test in [*2] by Arthur Gretton, 2007.

[*] https://www.mathworks.com/matlabcentral/fileexchange/25467-granger-causality-test

[*2] http://people.kyb.tuebingen.mpg.de/arthur/indep.htm

Acknowledgements: This code cannot be developed without the based-code by Chandler Lutz. I would like to thank Chandler Lutz, Arthur Gretton to provide the public code, and all comments in the matlab community.

[GfStatTest,follSimVal, notIndp,BICres,pval, F,c_v, threshIndp,testStatIndp] = DTW_granger_cause(x,y,alpha,max_lag,sigmaTHS)

INPUT: x,y is a time series of real numbers where x(t) is a real value at time t

INPUT: alpha, is the significance level of both f-test and independence test

INPUT: max_lag is a maximum, time lag and sigmaTHS is a following relation threshold

OUTPUT: GfStatTest = 1 if f-test of VR-Granger reject the H_0 that y does not VR-Granger cuases x with sig-level alpha, otherwise zero.

OUTPUT: follSimVal is between 0.1 to 1 if x follows y, follSimVal is between -1 to -0.1 if y follows l , and around zero if no following relation between x and y.

============

About

The framework of VL-Granger causality inference

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages