Skip to contents

Package website: release | dev

This package provides a common framework for optimization including

  • Optimizer: Objects of this class allow you to optimize an object of the class OptimInstance.
  • OptimInstance: Defines the optimization problem, consisting of an Objective, the search_space and a Terminator. All evaluations on the OptimInstance will be automatically stored in its own Archive.
  • Objective: Objects of this class contain the objective function. The class ensures that the objective function is called in the right way and defines, whether the function should be minimized or maximized.
  • Terminator: Objects of this class control the termination of the optimization independent of the optimizer.

Various optimization methods are already implemented e.g. grid search, random search and generalized simulated annealing.

Resources

Installation

Install the last release from CRAN:

Install the development version from GitHub:

remotes::install_github("mlr-org/bbotk")

Examples

Optimization

# define objective function
fun = function(xs) {
  - (xs[[1]] - 2)^2 - (xs[[2]] + 3)^2 + 10
}

# set domain
domain = ps(
  x1 = p_dbl(-10, 10),
  x2 = p_dbl(-5, 5)
)

# set codomain
codomain = ps(
  y = p_dbl(tags = "maximize")
)

# create Objective object
objective = ObjectiveRFun$new(
  fun = fun,
  domain = domain,
  codomain = codomain,
  properties = "deterministic"
)

# Define termination criterion
terminator = trm("evals", n_evals = 10)

# create optimization instance
instance = OptimInstanceSingleCrit$new(
  objective = objective,
  terminator = terminator
)

# load optimizer
optimizer = opt("gensa")

# trigger optimization
optimizer$optimize(instance)
##        x1        x2  x_domain        y
## 1: 2.0452 -2.064743 <list[2]> 9.123252
# best performing configuration
instance$result
##        x1        x2  x_domain        y
## 1: 2.0452 -2.064743 <list[2]> 9.123252
# all evaluated configuration
as.data.table(instance$archive)
##            x1        x2          y           timestamp batch_nr x_domain_x1
##  1: -4.689827 -1.278761 -37.716445 2022-09-10 18:22:05        1   -4.689827
##  2: -5.930364 -4.400474 -54.851999 2022-09-10 18:22:05        2   -5.930364
##  3:  7.170817 -1.519948 -18.927907 2022-09-10 18:22:05        3    7.170817
##  4:  2.045200 -1.519948   7.807403 2022-09-10 18:22:05        4    2.045200
##  5:  2.045200 -2.064742   9.123250 2022-09-10 18:22:05        5    2.045200
##  6:  2.045200 -2.064742   9.123250 2022-09-10 18:22:05        6    2.045200
##  7:  2.045201 -2.064742   9.123250 2022-09-10 18:22:05        7    2.045201
##  8:  2.045199 -2.064742   9.123250 2022-09-10 18:22:05        8    2.045199
##  9:  2.045200 -2.064741   9.123248 2022-09-10 18:22:05        9    2.045200
## 10:  2.045200 -2.064743   9.123252 2022-09-10 18:22:05       10    2.045200
##     x_domain_x2
##  1:   -1.278761
##  2:   -4.400474
##  3:   -1.519948
##  4:   -1.519948
##  5:   -2.064742
##  6:   -2.064742
##  7:   -2.064742
##  8:   -2.064742
##  9:   -2.064741
## 10:   -2.064743

Quick optimization with bb_optimize

library(bbotk)

# define objective function
fun = function(xs) {
  c(y1 = - (xs[[1]] - 2)^2 - (xs[[2]] + 3)^2 + 10)
}

# optimize function with random search
result = bb_optimize(fun, method = "random_search", lower = c(-10, -5), upper = c(10, 5),
  max_evals = 100)

# optimized parameters
result$par
##           x1       x2
## 1: -7.982537 4.273021
# optimal outcome
result$value
##        y1 
## -142.5479