R/newmcga.r

In mcga: Machine Coded Genetic Algorithms for Real-Valued Optimization Problems

# R File
# This is a wrapper file for the C library
mcga <-function (popsize, chsize, crossprob = 1.0, mutateprob =
		  0.01, elitism = 1, minval, maxval, maxiter = 10, evalFunc)
{

  par <-as.double(rep (0, chsize));
  best <-as.double(rep (0, chsize));
  population <-as.double(rep (0, chsize * popsize));
  costs <-as.double(rep (0, popsize));
  envv <-.GlobalEnv;
  result <-.Call ("mcga", popsize, chsize, crossprob, mutateprob, elitism,
		   minval, maxval, maxiter, par, best, evalFunc, population,
		   costs, envv, PACKAGE = "mcga");
  resmat <-matrix (population, ncol = chsize, nrow = popsize);
  return (list
	  (population =
	   matrix (population, popsize, chsize, byrow = TRUE), costs =
	   costs, popsize = popsize, chsize = chsize, crossprob =
	   crossprob, mutateprob = mutateprob, elitism = elitism, minval =
	   minval, maxval = maxval, maxiter = maxiter));
}


multi_mcga <-function (popsize, chsize, crossprob = 1.0, mutateprob =
                  0.01, elitism = 1, minval, maxval, maxiter = 10, numfunc, evalFunc)
{

  par <-as.double(rep (0, chsize));
  best <- as.double(rep (0, chsize));
  population <-as.double(rep (0, chsize * popsize));
  costs <- as.double(rep(0, popsize * numfunc));
  ranks <- as.double(rep (0, popsize));
  envv <-.GlobalEnv;
  result <-.Call ("multi_mcga", popsize, chsize, crossprob, mutateprob, elitism,
                   minval, maxval, maxiter, par, best, evalFunc, numfunc,  population,
                   costs, ranks, envv, PACKAGE = "mcga");
  resmat <-matrix (population, ncol = chsize, nrow = popsize);
  return (list
          (population =
           matrix (population, popsize, chsize, byrow = TRUE), costs =
           matrix (costs, popsize, numfunc, byrow=TRUE), ranks=ranks, popsize = popsize, chsize = chsize, crossprob =
           crossprob, mutateprob = mutateprob, elitism = elitism, minval =
           minval, maxval = maxval, maxiter = maxiter));
}