Mutate offspring with an adaptive rate

The mutation rate increases linearly with the generation index. Real-coded mode applies Gaussian perturbation; binary mode flips selection bits. When a gene network is supplied, the perturbation of each gene is damped by its relative connectivity (degree), so network hubs change less.

Usage

mutation_cpp(
  offspring,
  mutation_rate,
  iteration,
  max_iterations,
  network = NULL,
  representation = "real"
)

Arguments

offspring

Numeric matrix of offspring.

mutation_rate

Base probability of mutation.

iteration

Current GA iteration.

max_iterations

Maximum number of GA iterations.

network

Optional symmetric adjacency matrix (rows: genes, columns: genes).

representation

Either "real" or "binary".

Value

Numeric matrix of mutated offspring.

See also

initialize_population_cpp(), evaluate_fitness_cpp(), selection_cpp(), crossover_cpp(), replacement_cpp(), bioga_main_cpp()

Examples

genomic_data <- matrix(rnorm(100), nrow = 10, ncol = 10)
population <- BioGA::initialize_population_cpp(genomic_data, population_size = 5)
fitness <- BioGA::evaluate_fitness_cpp(genomic_data, population, c(1.0, 0.5))
parents <- BioGA::selection_cpp(population, fitness, num_parents = 2)
offspring <- BioGA::crossover_cpp(parents, offspring_size = 2)
BioGA::mutation_cpp(offspring, mutation_rate = 0.1, iteration = 1,
  max_iterations = 100)
#>            [,1]      [,2]      [,3]      [,4]     [,5]      [,6]       [,7]
#> [1,] -0.6002596 0.5539177 0.5846137 0.7799651 1.253815 0.3796395 -0.2950715
#> [2,] -0.6002596 0.5539177 0.5846137 0.7799651 1.253815 0.3796395 -0.3186994
#>             [,8]      [,9]    [,10]
#> [1,] -0.04287046 0.8781335 1.360652
#> [2,] -0.04287046 0.8781335 1.360652