A Library for Differential Evolution Optimization
Usage
A simple usage example:
import 'package:de/de.dart';
double f(List<double> x) {
return -(pow(x[0] - 3.14159265, 2.0) + pow(x[1] - exp(1.0), 2.0));
}
void main(){
DE de = DE(f, [-500, -500], [500, 500], cr: 0.99, F: 1.20, popsize: 300);
de.iterateN(5000);
Solution best = de.getBest();
print(best);
}Details
- The objective function is always in the form of
double f(List<double> x){
}-
The optimization is always a maximation. If the objective function of the problem is a minimization, the returned value of the objective function can be multiplied by -1.0.
-
The default parameters of the classical Differential Evolution Optimization are set to cr = 0.8 and F = 1.0. These optimization parameters can be customized in the constructor of DE class.