addIterations: Run Additional Optimization Iterations

In ParBayesianOptimization: Parallel Bayesian Optimization of Hyperparameters

Run Additional Optimization Iterations

Description

Use this function to continue optimization of a bayesOpt object.

Usage

addIterations(
  optObj,
  iters.n = 1,
  iters.k = 1,
  otherHalting = list(timeLimit = Inf, minUtility = 0),
  bounds = optObj$bounds,
  acq = optObj$optPars$acq,
  kappa = optObj$optPars$kappa,
  eps = optObj$optPars$eps,
  gsPoints = optObj$optPars$gsPoints,
  convThresh = optObj$optPars$convThresh,
  acqThresh = optObj$optPars$acqThresh,
  errorHandling = "stop",
  saveFile = optObj$saveFile,
  parallel = FALSE,
  plotProgress = FALSE,
  verbose = 1,
  ...
)

Arguments

optObj	an object of class bayesOpt.
iters.n	The total number of additional times to sample the scoring function.
iters.k	integer that specifies the number of times to sample FUN at each Epoch (optimization step). If running in parallel, good practice is to set iters.k to some multiple of the number of cores you have designated for this process. Must belower than, and preferrably some multiple of iters.n.
otherHalting	Same as bayesOpt()
bounds	Same as bayesOpt()
acq	Same as bayesOpt()
kappa	Same as bayesOpt()
eps	Same as bayesOpt()
gsPoints	Same as bayesOpt()
convThresh	Same as bayesOpt()
acqThresh	Same as bayesOpt()
errorHandling	Same as bayesOpt()
saveFile	Same as bayesOpt()
parallel	Same as bayesOpt()
plotProgress	Same as bayesOpt()
verbose	Same as bayesOpt()
...	Same as bayesOpt()

Details

By default, this function uses the original parameters used to create optObj, however the parameters (including the bounds) can be customized. If new bounds are used which cause some of the prior runs to fall outside of the bounds, these samples are removed from the optimization procedure, but will remain in scoreSummary. FUN should return the same elements and accept the same inputs as the original, or this function may fail.

Value

An object of class bayesOpt having run additional iterations.

Examples

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 = 1
  , gsPoints = 10
)
Results <- addIterations(Results,iters.n=1)

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