How do I use Genetic Algorithm?

Question

I am trying to learn how to use Genetic algorithm (GA) in r. So, here is my code:

library(genalg)

evalf <- function(x) {
      soln <- dnorm(x, mean = 16, sd = 4) 
      return(soln)
    }

iter = 100
GAmodel <- rbga.bin(size = 5, popSize = 200, iters = iter, mutationChance = 0.002, elitism = F, evalFunc = evalf)

return with the warning, There were 50 or more warnings (use warnings() to see the first 50) as follows:

1: In evalVals[object] = evalFunc(population[object, ]) : number of items to replace is not a multiple of replacement length

Further, I think GA does not return the correct answer since I got as follows:

cat(summary.rbga(GAmodel))

return,

GA Settings
  Type                  = binary chromosome
  Population size       = 200
  Number of Generations = 100
  Elitism               = FALSE
  Mutation Chance       = 0.002

Search Domain
  Var 1 = [,]
  Var 0 = [,]

GA Results
  Best Solution : 0 0 1 0 1 

If I convert the Best Solution using binary2decimal(c(0,0,1,0,1)), I got 5. However the correct answer should be 16.

Could you please give me suggestions on how to solve this problem using GA ?

Thank you

Answer 1

To solve this problem using GA you need to change evalf to:

evalf <- function(x) 1/dnorm(binary2decimal(x), mean = 16, sd = 4)

It seems that fitness function should return smaller value for data that is closer to solution. Also fitness function should return the value - not a vector - so you need to convert chromosome by using binary2decimal.