The GAREGEN subroutine replaces the current solution population by applying selection, crossover, and mutation for a genetic algorithm optimization problem. The input arguments to the GAREGEN call are as follows:

The GAREGEN call applies the genetic algorithm to create a new solution population from the current population. As the first step, if the elite parameter of the corresponding GASETSEL call is nonzero, the best elite members of the current population are copied into the new population, sorted by objective value with the best objective value first. If a crossover operator has been specified in a corresponding GASETCRO call or a default crossover operator is in effect, the remaining members of the population are determined by selecting members of the current population, applying the crossover operator to generate offspring, and mutating the offspring according to the mutation probability and mutation operator. Either the mutation probability and operator are specified in the corresponding GASETMUT call or, if no such call is made, a default value of 0.05 is assigned to the mutation probability and a default mutation operator is assigned based on the problem encoding (see the GASETMUT call). The offspring are then transferred to the new population. If the no-crossover option is specified in the GASETCRO call, then only mutation is applied to the non-elite members of the current population to form the new population. After the new population is formed, it becomes the current solution population, and the objective function specified in the GASETOBJ call is evaluated for each member.

See the GASETUP function for an example.