updatePool: Updating Chromosome Pool

In garylin2099/GA: Genetic Algorithm for Variable Selection in Regression

Description

updatePool is used to update the chromosome pool in each iteration of the genetic algorithms. It returns a vector of length P*C, which is loaded by all genes of next generation chromosomes.

Usage

updatePool(
  pool,
  fitness,
  tournamentSelection,
  groupNum,
  oneParentRandom,
  numCrossoverSplit,
  mutationRate,
  maxMutationRate,
  iterCounter,
  maxIter
)

Arguments

pool	a matrix of booleans representing the chromosome pools
fitness	a vector indicating the fitness score of each chromosome in the pool
tournamentSelection	a logical indicating whether to use tournament selection. If FALSE, rank-based selection is used.
groupNum	number of groups to be partitioned in each round of the tournament selection. Value must be between 2 and poolSize / 2 (inclusive) to ensure meaningful tournaments
oneParentRandom	a logical applying to rank-based selection. If TRUE, then we select one parent with the probability being its fitness score and the other randomly, otherwise, select both parents according to their fitness score.
numCrossoverSplit	number of crossover points.
mutationRate	constant mutation rate. Value must be between 0 and 1.
maxMutationRate	the mutation rate expected by the user when iteration limit is reached. If specified, then the mutation rate is linearly increasing in each iteration
iterCounter	an integer specifying which iteration in the the current update
maxIter	maximum number of iterations of updating the generation.

garylin2099/GA documentation built on Feb. 8, 2020, 1:13 a.m.