gena.population: Population

In gena: Genetic Algorithm and Particle Swarm Optimization

Population

Description

Initialize the population of chromosomes.

Usage

gena.population(pop.n, lower, upper, pop.initial = NULL, method = "uniform")

Arguments

pop.n	positive integer representing the number of chromosomes in population.
lower	numeric vector which i-th element determines the minimum possible value for i-th gene.
upper	numeric vector which i-th element determines the maximum possible value for i-th gene.
pop.initial	numeric matrix which rows are initial chromosomes suggested by user.
method	string representing the initialization method to be used. For a list of possible values see Details.

Details

If "method = uniform" then i-th gene of each chromosome is randomly (uniformly) chosen between lower[i] and upper[i] bounds. If "method = normal" then i-th gene is generated from a truncated normal distribution with mean (upper[i] + lower[i]) / 2 and standard deviation (upper[i] - lower[i]) / 6 where lower[i] and upper[i] are lower and upper truncation bounds correspondingly. If "method = hypersphere" then population is simulated uniformly from the hypersphere with center upper - lower and radius sqrt(sum((upper - lower) ^ 2)) via rhypersphere function setting type = "inside".

Value

This function returns a matrix which rows are chromosomes.

References

B. Kazimipour, X. Li, A. Qin (2014). A review of population initialization techniques for evolutionary algorithms. 2014 IEEE Congress on Evolutionary Computation, 2585-2592, <doi:10.1109/CEC.2014.6900618>.

Examples

set.seed(123)
gena.population(pop.n = 10,
                lower = c(-1, -2, -3),
                upper = c(1, 0, -1),
                pop.initial = rbind(c(0, -1, -2),
                                    c(0.1, -1.2, -2.3)),
                method = "normal")

gena documentation built on Aug. 15, 2022, 9:08 a.m.