kl.div {yest} | R Documentation |
Compute Kullback-Leibler divergence between two multivariate normal distributions.
kl.div(V0,mu0,V1,mu1=NA,inv=TRUE)
V0 |
(inverse) variance matrix of the first distribution |
mu0 |
expected value of the first distribution |
V1 |
(inverse) variance matrix of the second distribution |
mu1 |
expected value of the second distribution |
inv |
whether V0 and V1 are inverse varience (TRUE) or variance (FALSE) matrices |
If mu1
is NA
then a zero vector is taken as this parameter.
v1<-ind.rgauss(type=12) v2<-ind.rgauss(type=50) kl.div(v1,rep(0,4),v2)