Learn R Programming
miesmuschel (version 0.0.4-3)

mlr_terminators_budget: Terminator that Limits Total Budget Component Evaluation

Description

Terminator that terminates after the sum (or similar aggregate) of a given "budget" search space component croses a threshold.

Arguments

Dictionary

This Terminator can be created with the short access form trm() (trms() to get a list), or through the dictionary mlr_terminators in the following way:

# preferred
trm("budget")
trms("budget")  # takes vector IDs, returns list of Terminators
# long form
mlr_terminators$get("budget")

Configuration Parameters

  • budget :: numeric(1)
    Total budget available, after which to stop. Not initialized and should be set to the desired value during construction.

  • aggregate :: function
    Function taking a vector of values of the budget search space component, returning a scalar value to be compared to the budget configuration parameter. If this function returns a value greater or equal to budget the termination criterion is matched. Calling this function with NULL must return the lower bound of the budget value; percentage progress is reported as the progress from this lower bound to the value of budget. Initialized to sum().

Super class

bbotk::Terminator -> TerminatorBudget

Methods

Inherited methods

Method new()

Initialize the TerminatorBudget object.

Usage

TerminatorBudget$new()

Method is_terminated()

Is TRUE if when the termination criterion is matched, FALSE otherwise.

Usage

TerminatorBudget$is_terminated(archive)

Arguments

archive

Archive Archive to check.

Returns

logical(1): Whether to terminate.

Method clone()

The objects of this class are cloneable with this method.

Usage

TerminatorBudget$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

Examples

Run this code
library("bbotk")
# Evaluate until sum of budget component of evaluated configs is >= 100
trm("budget", budget = 100)

# Evaluate until sum of two to the power of budget component is >= 100
trm("budget", budget = 1024, aggregate = function(x) sum(2 ^ x))

Run the code above in your browser using DataLab