updateGP: Update Gaussian Processes in a bayesOpt Object

In ParBayesianOptimization: Parallel Bayesian Optimization of Hyperparameters

Update Gaussian Processes in a bayesOpt Object

Description

To save time, Gaussian processes are not updated after the last iteration in addIterations(). The user can do this manually, using this function if they wish. This is not necessary to continue optimization using addIterations.

Usage

updateGP(optObj, bounds = optObj$bounds, verbose = 1, ...)

Arguments

optObj	an object of class bayesOpt
bounds	The bounds to scale the parameters within.
verbose	Should the user be warned if the GP is already up to date?
...	passed to DiceKriging::km()

Value

An object of class bayesOpt with updated Gaussian processes.

Examples

# Create initial object
scoringFunction <- function(x) {
  a <- exp(-(2-x)^2)*1.5
  b <- exp(-(4-x)^2)*2
  c <- exp(-(6-x)^2)*1
  return(list(Score = a+b+c))
}

bounds <- list(x = c(0,8))

Results <- bayesOpt(
    FUN = scoringFunction
  , bounds = bounds
  , initPoints = 3
  , iters.n = 2
  , gsPoints = 10
)

# At this point, the Gaussian Process has not been updated
# with the most recent results. We can update it manually:
Results <- updateGP(Results)

ParBayesianOptimization documentation built on Oct. 18, 2022, 5:07 p.m.