differenciation and Box-Cox transformation
[yt]=transdif(y,lambda,d,ds,s)
* y = (nxk) data matrix
* lambda = a scalar, parameter of the Box-Cox transformation
* d = a scalar, the number of regular differences (1-B)^d
* ds = (Sx1) matrix containing the number of seasonal differences
* s = (rx1) matrix containing the seasonal periods
* yt = (n-d-sum(ds*s))xk matrix of transformed data
load(GROCERDIR+'/data/varma_d.dat'); elec_cons = transdif(seriesa,0,1,1,12); // Example taken from varma function varma_d1(). Elec_cons is equal to (1-L)(1-L^12)log(seriesa). s = transdif(seriesa,-1,0,0,12); //returns the box-cox transformation of seriesa with lambda = -1 (that is 1-1/seriesa) | ![]() | ![]() |