Stochastic parameter estimation
Solution 1:
MLE() takes two variables, yet you gave the optim() function three parameters. Essentially, the optim() function expects b in your MLE function to be a vector of two spots. If you wanted to optimize b and y, for example, this will work.
MLE <- function(b){
J <<- vector(length = Ti)
for(i in 1:Ti){
J[i] <<- log(PL(i, b[1], b[2]))
}
return(sum(J))
}
MLE(c(1, 0.5))
optim(c(1, 1), MLE)
Now b is b[1] and y is b[2]. I'm not sure if that's what you wanted to optimize, though.