/// Returns the original version of the decision variables ready to be fed
/// to the original problem
decision_vector rotated::derotate(const decision_vector& x_normed) const
{
// This may be outside of the original domain, due to the
// relaxed variable bounds after rotation -- project it back if so.
// 1. De-rotate the vector in the normalized space
Eigen::VectorXd x_normed_vec = Eigen::VectorXd::Zero(x_normed.size());
Eigen::VectorXd x_derotated_vec;
for(base::size_type i = 0; i < x_normed.size(); i++){
x_normed_vec(i) = x_normed[i];
}
x_derotated_vec = m_InvRotate * x_normed_vec;
// 2. De-normalize the de-rotated vector to the original bounds
decision_vector x_derotated(x_normed.size(), 0);
for(base::size_type i = 0; i < x_normed.size(); i++){
x_derotated[i] = x_derotated_vec(i);
}
decision_vector x_wild = denormalize_to_original(x_derotated);
// 3. The de-normalized vector may be out of bounds, if so project back in
decision_vector x = projection_via_clipping(x_wild);
return x;
}
/// Apply noise on the decision vector based on rho
void robust::inject_noise_x(decision_vector &x) const
{
// We follow the algorithm at
// http://math.stackexchange.com/questions/87230/picking-random-points-in-the-volume-of-sphere-with-uniform-probability
// 0. Define the radius
double radius = m_rho * pow(m_uniform_dist(m_drng),1.0/x.size());
// 1. Sampling N(0,1) on each dimension
std::vector<double> perturbation(x.size(), 0.0);
double c2=0;
for(size_type i = 0; i < perturbation.size(); i++){
perturbation[i] = m_normal_dist(m_drng);
c2 += perturbation[i]*perturbation[i];
}
// 2. Normalize the vector
for(size_type i = 0; i < perturbation.size(); i++){
perturbation[i] *= (radius / sqrt(c2) );
x[i] += perturbation[i];
}
// 3. Clip the variables to the valid bounds
for(base::size_type i = 0; i < x.size(); i++){
x[i] = std::max(x[i], get_lb()[i]);
x[i] = std::min(x[i], get_ub()[i]);
}
}
// Used to normalize the original upper and lower bounds
// to [-1, 1], at each dimension
decision_vector rotated::normalize_to_center(const decision_vector& x) const
{
decision_vector normalized_x(x.size(), 0);
for(base::size_type i = 0; i < x.size(); i++) {
if (m_normalize_scale[i] == 0) { //If the bounds witdth is zero
normalized_x[i] = 0;
}
else {
normalized_x[i] = (x[i] - m_normalize_translation[i]) / m_normalize_scale[i];
}
}
return normalized_x;
}
double antibodies_problem::compute_distance(const decision_vector &x) const {
double distance = 0.;
// hamming distance
switch(m_method) {
case(algorithm::cstrs_immune_system::HAMMING): {
const decision_vector &lb = get_lb();
const decision_vector &ub = get_ub();
for(decision_vector::size_type i=0; i<x.size(); i++) {
std::vector<int> current_binary_gene = double_to_binary(x.at(i), lb.at(i), ub.at(i));
for(decision_vector::size_type j=0; j<m_pop_antigens.size(); j++) {
std::vector<int> antigens_binary_gene = double_to_binary((m_pop_antigens.at(j)).at(i), lb.at(i), ub.at(i));
for(std::vector<int>::size_type k=0; k<antigens_binary_gene.size(); k++) {
distance += antigens_binary_gene.at(k) && current_binary_gene.at(k);
}
}
}
// we take the inverse of the distance as the measure we need is
// how close the x is from the antigen population
// which means that we need to maximize the ressemblance
distance = - distance;
break;
}
case(algorithm::cstrs_immune_system::EUCLIDEAN): {
for(decision_vector::size_type j=0; j<m_pop_antigens.size(); j++) {
double euclid = 0.;
const decision_vector &antigen_decision = m_pop_antigens.at(j);
for(decision_vector::size_type i=0; i<x.size(); i++) {
euclid += std::pow(x.at(i) - antigen_decision.at(i),2);
}
distance += std::sqrt(euclid);
}
break;
}
}
return distance;
}
// Write into retval the gradient of the continuous part of the objective function of prob calculated in input.
void gsl_gradient::objfun_numdiff_central(gsl_vector *retval, const problem::base &prob, const decision_vector &input, const double &step_size)
{
if (input.size() != prob.get_dimension()) {
pagmo_throw(value_error,"invalid input vector dimension in numerical differentiation of the objective function");
}
if (prob.get_f_dimension() != 1) {
pagmo_throw(value_error,"numerical differentiation of the objective function cannot work on multi-objective problems");
}
// Size of the continuous part of the problem.
const problem::base::size_type cont_size = prob.get_dimension() - prob.get_i_dimension();
// Structure to pass data to the wrapper.
objfun_numdiff_wrapper_params pars;
pars.x = input;
pars.f.resize(1);
pars.prob = &prob;
// GSL function.
gsl_function F;
F.function = &objfun_numdiff_wrapper;
F.params = (void *)&pars;
double result, abserr;
// Numerical differentiation component by component.
for (problem::base::size_type i = 0; i < cont_size; ++i) {
pars.coord = i;
gsl_deriv_central(&F,input[i],step_size,&result,&abserr);
gsl_vector_set(retval,i,result);
}
}
/**
* @return the encoded string (std::string)
*/
std::string string_match::pretty(const decision_vector &x) const {
std::string retval;
for (decision_vector::size_type i = 0; i < x.size(); ++i) {
retval += char(x[i]);
}
return retval;
}
/// Implementation of the objective function.
void lavor_maculan::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
const decision_vector::size_type n = x.size();
f[0] = 0.0;
for (decision_vector::size_type i = 0 ; i < n ; ++i) {
f[0] += 1 + cos(3 * x[i]) + (((i % 2 == 1) ? 1 : -1) / (sqrt(10.60099896 - 4.141720682 * cos(x[i]))));
}
}
/// Implementation of the objective function.
void rosenbrock::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
const decision_vector::size_type n = x.size();
f[0]=0;
for (decision_vector::size_type i=0; i<n-1; ++i){
f[0] += 100 * (x[i]*x[i] -x[i+1])*(x[i]*x[i] -x[i+1]) + (x[i]-1)*(x[i]-1);
}
}
// For a decision_vector in the normalized [-1, 1] space,
// perform the inverse operation of normalization to get it's
// location in the original space
decision_vector rotated::denormalize_to_original(const decision_vector& x_normed) const
{
decision_vector denormalized_x(x_normed.size(), 0);
for(base::size_type i = 0; i < x_normed.size(); i++){
denormalized_x[i] = (x_normed[i] * m_normalize_scale[i]) + m_normalize_translation[i];
}
return denormalized_x;
}
/// Implementation of the constraint function.
void luksan_vlcek_3::compute_constraints_impl(constraint_vector &c, const decision_vector &x) const
{
int n = x.size();
c[0] = 3.*std::pow(x[0],3) + 2.*x[1] - 5. + std::sin(x[0]-x[1])*std::sin(x[0]+x[1]) - m_cub[0];
c[1] = m_clb[0] - ( 3.*std::pow(x[0],3) + 2.*x[1] - 5. + std::sin(x[0]-x[1])*std::sin(x[0]+x[1]) );
c[2] = 4.*x[n-3] - x[n-4]*std::exp(x[n-4]-x[n-3]) - 3 - m_cub[1];
c[3] = m_clb[1] - ( 4.*x[n-3] - x[n-4]*std::exp(x[n-4]-x[n-3]) - 3 );
}
/// Implementation of the objective function.
void sample_return::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
//We split the decision vector in the two legs
std::copy(x.begin(),x.begin()+6,x_leg1.begin());
std::copy(x.begin()+6,x.begin()+12,x_leg2.begin());
x_leg1[4] = x_leg1[4] * m_Tmax;
x_leg2[4] = (m_Tmax - x_leg1[4] - x_leg2[0]) * x_leg2[4];
//We account for the waiting time
x_leg2[0] += x_leg1[0] + x_leg1[4];
double dummy = 0;
MGA_DSM(x_leg1, m_leg1, dummy);
MGA_DSM(x_leg2, m_leg2, dummy);
f[0] = m_leg1.DV[0] + m_leg1.DV[1] + m_leg1.DV[2] +
m_leg2.DV[0] + m_leg2.DV[1] + std::max(0.0,m_leg2.DV[2] - 5.5);
}
/// Implementation of the constraint function.
void earth_planet::compute_constraints_impl(constraint_vector &c, const decision_vector &x) const
{
// We decode the decision vector into a multiple fly-by trajectory
trajectory.init_from_full_vector(x.begin(),x.end(),encoding);
// We evaluate the state mismatch at the mid-point. And we use astronomical units to scale them
trajectory.evaluate_all_mismatch_con(c.begin(), c.begin() + 7);
for (int i=0; i<3; ++i) c[i]/=ASTRO_AU;
for (int i=3; i<6; ++i) c[i]/=ASTRO_EARTH_VELOCITY;
// We evaluate the constraints on the throttles writing on the 7th mismatch constrant (mass is off)
trajectory.get_leg(0).get_throttles_con(c.begin() + 6, c.begin() + 6 + n_segments);
// We evaluate the constraint on the initial launch velocity
c[6 + n_segments] = (trajectory.evaluate_leg_vinf2_i(0) - vmax*vmax) / ASTRO_EARTH_VELOCITY / ASTRO_EARTH_VELOCITY;
// We evaluate the linear constraint on the epochs (tf > ti)
c[7 + n_segments] = trajectory.get_leg(0).get_t_i().mjd2000() - trajectory.get_leg(0).get_t_f().mjd2000();
}
/// Implementation of the objective function.
void rastrigin::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 1);
const double omega = 2.0 * boost::math::constants::pi<double>();
f[0] = 0;
const decision_vector::size_type n = x.size();
for (decision_vector::size_type i = 0; i < n; ++i) {
f[0] += x[i] * x[i] - 10.0 * std::cos(omega * x[i]);
}
f[0] += 10.0 * n;
}
/// Implementation of the objective function.
void michalewicz::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 1);
decision_vector::size_type n = x.size();
double retval = 0.0;
for (decision_vector::size_type i=0; i<n; i++){
retval -= sin(x[i]) * pow(sin((i+1)*x[i]*x[i]/boost::math::constants::pi<double>()) , 2*m_m);
}
f[0] = retval;
}
/// Implementation of the objective function.
void schwefel::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 1);
std::vector<double>::size_type n = x.size();
double value=0;
for (std::vector<double>::size_type i=0; i<n; i++){
value += x[i] * sin(sqrt(fabs(x[i])));
}
f[0] = 418.9828872724338 * n - value;
}
/// Implementation of the objective function.
void dejong::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 1);
decision_vector::size_type n = x.size();
double retval = 0.0;
for (decision_vector::size_type i=0; i<n; i++){
retval += x[i]*x[i];
}
f[0] = retval;
}
/// Implementation of the objective function.
void luksan_vlcek_3::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
f[0] = 0.;
for (decision_vector::size_type i=0; i<(x.size()-2)/2; i++)
{
double a1 = x[2*i]+10.*x[2*i+1];
double a2 = x[2*i+2] - x[2*i+3];
double a3 = x[2*i+1] - 2.*x[2*i+2];
double a4 = x[2*i] - x[2*i+3];
f[0] += a1*a1 + 5.*a2*a2 + std::pow(a3,4)+ 10.*std::pow(a4,4);
}
}
std::string gtoc5_rendezvous::pretty(const decision_vector &x) const
{
using namespace kep_toolbox;
// We set the leg.
const epoch epoch_i(x[0],epoch::MJD), epoch_f(x[1] + x[0],epoch::MJD);
array3D v0, r0, vf, rf;
m_source.eph(epoch_i,r0,v0);
m_target.eph(epoch_f,rf,vf);
m_leg.set_leg(epoch_i,sc_state(r0,v0,m_leg.get_spacecraft().get_mass()),x.begin() + 3, x.end(),epoch_f,sc_state(rf,vf,x[2]),ASTRO_MU_SUN);
std::ostringstream oss;
oss << m_leg << '\n' << m_source << '\n' << m_target << '\n';
return oss.str();
}
/// Implementation of the objective function.
void zdt2::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 2);
pagmo_assert(x.size() == 30);
double g = 0;
f[0] = x[0];
for(problem::base::size_type i = 2; i < 30; ++i) {
g += x[i];
}
g = 1 + (9 * g) / 29;
f[1] = g * ( 1 - (x[0]/g)*(x[0]/g));
}
/// Implementation of the objective function.
void zdt6::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 2);
pagmo_assert(x.size() == 10);
double g = 0;
f[0] = 1 - exp(-4*x[0])*pow(sin(6*m_pi*x[0]),6);
for(problem::base::size_type i = 2; i < 10; ++i) {
g += x[i];
}
g = 1 + (9 * g) / 9;
f[1] = g * ( 1 - (f[0]/g)*(f[0]/g));
}
/// Implementation of the constraint function.
void gtoc5_rendezvous::compute_constraints_impl(constraint_vector &c, const decision_vector &x) const
{
using namespace kep_toolbox;
// We set the leg.
const epoch epoch_i(x[0],epoch::MJD), epoch_f(x[1] + x[0],epoch::MJD);
array3D v0, r0, vf, rf;
m_source.eph(epoch_i,r0,v0);
m_target.eph(epoch_f,rf,vf);
m_leg.set_leg(epoch_i,sc_state(r0,v0,m_leg.get_spacecraft().get_mass()),x.begin() + 3, x.end(),epoch_f,sc_state(rf,vf,x[2]),ASTRO_MU_SUN);
// We evaluate the state mismatch at the mid-point. And we use astronomical units to scale them
m_leg.get_mismatch_con(c.begin(), c.begin() + 7);
for (int i=0; i<3; ++i) c[i]/=ASTRO_AU;
for (int i=3; i<6; ++i) c[i]/=ASTRO_EARTH_VELOCITY;
c[6] /= m_leg.get_spacecraft().get_mass();
// We evaluate the constraints on the throttles writing on the 7th mismatch constrant (mass is off)
m_leg.get_throttles_con(c.begin() + 7, c.begin() + 7 + m_n_segments);
}
std::string gtoc5_launch::pretty(const decision_vector &x) const
{
using namespace kep_toolbox;
// 1 - We set the leg.
const epoch epoch_i(x[0],epoch::MJD), epoch_f(x[1] + x[0],epoch::MJD);
array3D v0, r0, vf, rf;
m_earth.get_eph(epoch_i,r0,v0);
m_target.get_eph(epoch_f,rf,vf);
v0[0] += x[2];
v0[1] += x[3];
v0[2] += x[4];
m_leg.set_leg(epoch_i,sc_state(r0,v0,m_leg.get_spacecraft().get_mass()),x.begin() + 6, x.end(),epoch_f,sc_state(rf,vf,x[5]),ASTRO_MU_SUN);
std::ostringstream oss;
oss << m_leg << '\n' << m_earth << '\n' << m_target << '\n';
return oss.str();
}
/// Implementation of the objective function.
void levy5::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 1);
decision_vector::size_type n = x.size();
double isum = 0.0;
double jsum = 0.0;
f[0] = 0;
for ( decision_vector::size_type j=0; j<n; j+=2 ) {
for ( int i=1; i<=5; i++ ) {
isum += (double)(i) * cos((double)(i-1)*x[j] + (double)(i));
jsum += (double)(i) * cos((double)(i+1)*x[j+1] + (double)(i));
}
}
f[0] = isum*jsum;
for ( decision_vector::size_type j=0; j<n; j+=2 )
f[0] += pow(x[j] + 1.42513,2) + pow(x[j+1] + 0.80032,2);
}
/// Implementation of the objective function.
void earth_planet::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
trajectory.init_from_full_vector(x.begin(),x.end(),encoding);
f[0] = trajectory.get_leg(0).evaluate_dv() / 1000;
}
/// Implementation of the objective function.
void sch::objfun_impl(fitness_vector &f, const decision_vector &x) const
{
pagmo_assert(f.size() == 2 && x.size() == 1);
f[0] = x[0]*x[0];
f[1] = (x[0]-2) * (x[0]-2);
}
void tsp_vrplc::compute_constraints_impl(constraint_vector &c, const decision_vector& x) const
{
decision_vector::size_type n_cities = get_n_cities();
switch( get_encoding() )
{
case FULL:
{
// 1 - We set the equality constraints
for (size_t i = 0; i < n_cities; i++) {
c[i] = 0;
c[i+n_cities] = 0;
for (size_t j = 0; j < n_cities; j++)
{
if(i==j) continue; // ignoring main diagonal
decision_vector::size_type rows = compute_idx(i, j, n_cities);
decision_vector::size_type cols = compute_idx(j, i, n_cities);
c[i] += x[rows];
c[i+n_cities] += x[cols];
}
c[i] = c[i]-1;
c[i+n_cities] = c[i+n_cities]-1;
}
//2 - We set the inequality constraints
//2.1 - First we compute the uj (see http://en.wikipedia.org/wiki/Travelling_salesman_problem#Integer_linear_programming_formulation)
// we start always out tour from the first city, without loosing generality
size_t next_city = 0,current_city = 0;
std::vector<int> u(n_cities);
for (size_t i = 0; i < n_cities; i++)
{
u[current_city] = i+1;
for (size_t j = 0; j < n_cities; j++)
{
if (current_city==j) continue;
if (x[compute_idx(current_city, j, n_cities)] == 1)
{
next_city = j;
break;
}
}
current_city = next_city;
}
int count=0;
for (size_t i = 1; i < n_cities; i++) {
for (size_t j = 1; j < n_cities; j++)
{
if (i==j) continue;
c[2*n_cities+count] = u[i]-u[j] + (n_cities+1) * x[compute_idx(i, j, n_cities)] - n_cities;
count++;
}
}
break;
}
case RANDOMKEYS:
break;
case CITIES:
{
std::vector<population::size_type> range(n_cities);
for (std::vector<population::size_type>::size_type i=0; i<range.size(); ++i)
{
range[i]=i;
}
c[0] = !std::is_permutation(x.begin(),x.end(),range.begin());
break;
}
}
return;
}
