如何获取整数形式的 GA 结果？

Question

我想在 R 中做一些优化 - 问题是结果应该是整数（1, 2 ...）而不是浮点数（41.7313418341801 41.0267840987071 ...等等）。是否可以编辑代码以找到总值？

代码：

library(genalg)
evaluate <- function(X=c()) {

a <- c(3400, 2400, 2400, 1400, 1200, 3400, 1300, 2500)
b <- c(5600, 4600, 5100, 2400, 2200, 3400, 1300, 2500)

returnVal = NA;
d <- sum(ifelse(X >= 4, a, b) * X)

if (d<150000) {
returnVal = sum(c(-200000,-130000,-145000,-75000, -55000, -210000, -150000, -20000)\*X);
} else {
returnVal =   10000000000000
}
round(returnVal)
}
rbga.results <- rbga(c(0,0,0,0,0,0,0,0), c(44,63,63,107,125,44,115,60),
popSize = 3000, evalFunc = evaluate, iters = 15,
verbose = TRUE, mutationChance = 0.01)
summary(rbga.results,echo=TRUE)

我明白了： 大会结果 最佳解决方案：41.7313418341801 41.0267840987071 44.4433741739485 31.3764953238424 69.9314918310847 21.4769318029284 11.3982368202414 0.266784215345979

但我希望收到的结果是： 41 41 44 ....等等（没有结果与假设不符的风险 - 成本应低于 150000）。

Answer 1

第

X

轮和人口结果：

library(genalg)
set.seed(1073902509L)

evaluate <- function(X=c()) {
  X <- round(X)
  a <- c(3400, 2400, 2400, 1400, 1200, 3400, 1300, 2500)
  b <- c(5600, 4600, 5100, 2400, 2200, 3400, 1300, 2500)
  d <- sum(ifelse(X >= 4, a, b) * X)
  
  if (d<150000) {
    sum(c(-200000,-130000,-145000,-75000, -55000, -210000, -150000, -20000)*X)
  } else {
    d*1e6
  }
}
system.time(
  rbga.results <- rbga(c(0,0,0,0,0,0,0,0), c(44,63,63,107,125,44,115,60),
                       popSize = 3000, evalFunc = evaluate, iters = 15,
                       verbose = FALSE, mutationChance = 0.01)
)
#>    user  system elapsed 
#>   10.89    0.69   11.60
rbga.results$population <- round(rbga.results$population)
summary(rbga.results,echo=TRUE)
#> GA Settings
#>   Type                  = floats chromosome
#>   Population size       = 3000
#>   Number of Generations = 15
#>   Elitism               = 600
#>   Mutation Chance       = 0.01
#> 
#> Search Domain
#>   Var 1 = [0,44]
#>   Var 2 = [0,63]
#>   Var 3 = [0,63]
#>   Var 4 = [0,107]
#>   Var 5 = [0,125]
#>   Var 6 = [0,44]
#>   Var 7 = [0,115]
#>   Var 8 = [0,60]
#> 
#> GA Results
#>   Best Solution : 4 5 5 13 0 4 59 0
rbga.results$population[which.min(rbga.results$evaluations),]
#> [1]  4  5  5 13  0  4 59  0
rbga.results$best
#>  [1]  -7355000  -7355000  -7355000  -9565000  -9565000 -11245000 -11245000
#>  [8] -11630000 -11630000 -11630000 -11630000 -12130000 -12130000 -12130000
#> [15] -12840000
evaluate(rbga.results$population[which.min(rbga.results$evaluations),])
#> [1] -12840000