Dear Users,
I would like to plot the log-likelihood (depending on two
parameters).
I wrote the following code:
library(mvtnorm)
sigma<-
matrix(c(4,2,2,3),ncol=2)
x<-rmvnorm(n=500,mean=c(1,2),sigma=sigma)
likel<-
function(param,data){
pos1<-which(data[,1]<0)
pos2<-which(data[,1]>=0)
#density
dens<-rep(0,nrow(data))
dens[pos1]<-dmvnorm(data[pos1,])
*param[1]
dens[pos2]<-dmvnorm(data[pos2,])*param[2]
#return log-
likelihood
sum(log(dens))
}
#plotting the log-likelihood
param1<-
seq(from=0,to=1,length=200)
param2<-seq(from=0,to=1,length=200)
combin<-
as.matrix(expand.grid(param1,param2,KEEP.OUT.ATTRS=FALSE))
likel.values<-
apply(combin,1,likel,data=x)
matr.likel.values<-matrix(likel.values,200,200)
persp(param1,param2,matr.likel.values,
xlab="param1",ylab="param2",
zlab="Likelihood",
theta=5,phi=20,expand=0.5,
col="lightblue",
ltheta=120,shade=0.75,ticktype="detailed")
It is all ok, but I am searching
for a new "lighting" way in order to earn time.
I need an alternative function
for apply(), when I calculate the log-likelihood for every single couple of
values (param1,param2).
Do you have any ideas?
Thank you in advance,
Enrico Foscolo
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
