/**
* Run the CORE algorithm
*
* @param[in,out] pop input/output pagmo::population to be evolved.
*/
void cstrs_core::evolve(population &pop) const
{
// store useful variables
const problem::base &prob = pop.problem();
const population::size_type pop_size = pop.size();
const problem::base::size_type prob_dimension = prob.get_dimension();
// get the constraints dimension
problem::base::c_size_type prob_c_dimension = prob.get_c_dimension();
//We perform some checks to determine wether the problem/population are suitable for CORE
if(prob_c_dimension < 1) {
pagmo_throw(value_error,"The problem is not constrained and CORE is not suitable to solve it");
}
if(prob.get_f_dimension() != 1) {
pagmo_throw(value_error,"The problem is multiobjective and CORE is not suitable to solve it");
}
// Get out if there is nothing to do.
if(pop_size == 0) {
return;
}
// generates the unconstrained problem
problem::con2uncon prob_unconstrained(prob);
// associates the population to this problem
population pop_uncon(prob_unconstrained);
// fill this unconstrained population
pop_uncon.clear();
for(population::size_type i=0; i<pop_size; i++) {
pop_uncon.push_back(pop.get_individual(i).cur_x);
}
// vector containing the infeasibles positions
std::vector<population::size_type> pop_infeasibles;
// Main CORE loop
for(int k=0; k<m_gen; k++) {
if(k%m_repair_frequency == 0) {
pop_infeasibles.clear();
// get the infeasible individuals
for(population::size_type i=0; i<pop_size; i++) {
if(!prob.feasibility_c(pop.get_individual(i).cur_c)) {
pop_infeasibles.push_back(i);
}
}
// random shuffle of infeasibles?
population::size_type number_of_repair = (population::size_type)(m_repair_ratio * pop_infeasibles.size());
// repair the infeasible individuals
for(population::size_type i=0; i<number_of_repair; i++) {
const population::size_type ¤t_individual_idx = pop_infeasibles.at(i);
pop.repair(current_individual_idx, m_repair_algo);
}
// the population is repaired, it can be now used in the new unconstrained population
// only the repaired individuals are put back in the population
for(population::size_type i=0; i<number_of_repair; i++) {
population::size_type current_individual_idx = pop_infeasibles.at(i);
pop_uncon.set_x(current_individual_idx, pop.get_individual(current_individual_idx).cur_x);
}
}
m_original_algo->evolve(pop_uncon);
// push back the population in the main problem
pop.clear();
for(population::size_type i=0; i<pop_size; i++) {
pop.push_back(pop_uncon.get_individual(i).cur_x);
}
// Check the exit conditions (every 40 generations, just as DE)
if(k % 40 == 0) {
decision_vector tmp(prob_dimension);
double dx = 0;
for(decision_vector::size_type i=0; i<prob_dimension; i++) {
tmp[i] = pop.get_individual(pop.get_worst_idx()).best_x[i] - pop.get_individual(pop.get_best_idx()).best_x[i];
dx += std::fabs(tmp[i]);
}
if(dx < m_xtol ) {
if (m_screen_output) {
std::cout << "Exit condition -- xtol < " << m_xtol << std::endl;
}
break;
}
double mah = std::fabs(pop.get_individual(pop.get_worst_idx()).best_f[0] - pop.get_individual(pop.get_best_idx()).best_f[0]);
if(mah < m_ftol) {
if(m_screen_output) {
std::cout << "Exit condition -- ftol < " << m_ftol << std::endl;
}
break;
}
// outputs current values
if(m_screen_output) {
std::cout << "Generation " << k << " ***" << std::endl;
std::cout << " Best global fitness: " << pop.champion().f << std::endl;
std::cout << " xtol: " << dx << ", ftol: " << mah << std::endl;
std::cout << " xtol: " << dx << ", ftol: " << mah << std::endl;
}
}
}
}
