Closed MatthiasPucher closed 4 years ago
I apologise for the huge delay, but I had to temporally be offline for some time, but now I'm back.
The answer to your question is to use the par
argument. From '?hydroPSO`:
"If the user provides ‘m’ parameter sets for ‘par’, they are used to overwrite the first ‘m’ parameter sets randomly defined according to the value of ‘Xini.type’"
D <- 5
lower <- rep(-5.12, D)
upper <- rep(5.12, D)
# initial parameter values:
x0 <- rep(0, D)
hydroPSO(fn=rastrigin, par=x0, lower=lower, upper=upper, control=list(write2disk=FALSE) )
You are not restricted to provide a single initial parameter set, but you might provide as many as the number of particles, where each row in par
represents a different parameter set and each column in par
represent a different parameter, in the same order you enter them in lower
and upper
Thanks, its highly appreciated. I will try this!
So far, I have not found a way to suggest the positions of particles. Would this make sense anyway? I intended to start particles at points, where I assume a good fit close by and by that shorten the calculation time. Would it then be possible, to suggest some positions and generate others similar to a default run?