songsenand
/
cgal
mirror of https://github.com/CGAL/cgal
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Don't recompute the weight every time

This commit is contained in:
Mael Rouxel-Labbé 2020-03-24 10:51:10 +01:00
parent f2303df4df
commit 25ac723c65
1 changed files with 3 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
Optimal_bounding_box/include/CGAL/Optimal_bounding_box/internal/evolution.h
View File

@ -75,7 +75,7 @@ public:
std::generate(group4.begin(), group4.end(), [&]{ return m_rng.get_int(0, im); }); std::generate(group4.begin(), group4.end(), [&]{ return m_rng.get_int(0, im); });
// crossover I, pick A or B // crossover I, pick A or B
FT bias = FT(0.1); constexpr FT lweight = 0.4, uweight = 0.6;
std::vector<Simplex> new_simplices(m); std::vector<Simplex> new_simplices(m);
@ -87,7 +87,7 @@ public:
const FT r{m_rng.get_double()}; const FT r{m_rng.get_double()};
const FT fitnessA = m_population[group1[i]][j].fitness_value(); const FT fitnessA = m_population[group1[i]][j].fitness_value();
const FT fitnessB = m_population[group2[i]][j].fitness_value(); const FT fitnessB = m_population[group2[i]][j].fitness_value();
const FT threshold = (fitnessA < fitnessB) ? (0.5 + bias) : (0.5 - bias); const FT threshold = (fitnessA < fitnessB) ? uweight : lweight;
if(r < threshold) if(r < threshold)
offspring[j] = m_population[group1[i]][j]; offspring[j] = m_population[group1[i]][j];
@ -99,8 +99,6 @@ public:
} }
// crossover II, combine information from A and B // crossover II, combine information from A and B
bias = 0.1; // @fixme should the bias change? What should be the initial value?
for(std::size_t i=0; i<second_group_size; ++i) for(std::size_t i=0; i<second_group_size; ++i)
{ {
Simplex offspring; Simplex offspring;
@ -108,7 +106,7 @@ public:
{ {
const FT fitnessA = m_population[group3[i]][j].fitness_value(); const FT fitnessA = m_population[group3[i]][j].fitness_value();
const FT fitnessB = m_population[group4[i]][j].fitness_value(); const FT fitnessB = m_population[group4[i]][j].fitness_value();
const FT lambda = (fitnessA < fitnessB) ? (0.5 + bias) : (0.5 - bias); const FT lambda = (fitnessA < fitnessB) ? uweight : lweight;
const FT rambda = 1 - lambda; // because the 'l' in 'lambda' stands for left const FT rambda = 1 - lambda; // because the 'l' in 'lambda' stands for left
const Matrix& lm = m_population[group3[i]][j].matrix(); const Matrix& lm = m_population[group3[i]][j].matrix();