import java.util.Random;

public class GAIndividual {

	private static Random randg = new Random(); // Random Generator
	// What if we choose a static random generator?
	// A unique random generator for all individuals!?

	public int genome_size;
	public float[] genome;
	public float fitness;

	public GAIndividual(int gsize,float[] min_range,float[] max_range){
	// Create a random individual of size gsize
	// i'th gene must be in range min_range[i] to max_range[i]
		genome_size = gsize;
		genome = new float[genome_size];
		for(int i=0; i < genome_size; i++){
			genome[i] = randg.nextFloat() * (max_range[i] - min_range[i]) + min_range[i];
		}
		evalFitness(); // evaluate fitness of this new individual
	}
	public GAIndividual(float d[]){
	// Create an individual that its genome is same as d[]
	// It is assumed min_range[i] <= d[i] <= max_range[i]
		genome_size = d.length;
		genome = new float[genome_size];
		for(int i=0; i < genome_size; i++){
			genome[i] = d[i];
		}
		evalFitness(); // evaluate fitness of this new individual
	}

	public GAIndividual mutate(float[] min_range,float[] max_range){
		// rate is the chance of each gene to be mutated
		// it is a good choice to set rate to 1/genome_size
		float rate = 1.0f / (float) genome_size;
		float[] result = new float[genome_size];
		for(int i=0; i < genome_size; i++)
			result[i] = genome[i];

		if( randg.nextBoolean() ){ // apply normal mutation
			for(int i=0; i < genome_size; i++)
				if( (float)Math.random() < rate)
					result[i] = randg.nextFloat() * (max_range[i] - min_range[i]) + min_range[i];
		}else{ // apply creep mutation
			for(int i=0; i < genome_size; i++)
				if( (float)Math.random() < rate)
					result[i] = result[i] + randg.nextFloat() * .2f - .1f;
		}

		return new GAIndividual(result);
	}

	public GAIndividual xover1p(GAIndividual m){
	// 1-point cross over
		int xpoint = 1 + randg.nextInt(genome_size - 1);
		float[] child = new float[genome_size];
		for(int i=0; i < xpoint; i++){
			child[i] = genome[i];
		}
		for(int i=xpoint; i < genome_size; i++){
			child[i] = m.genome[i];
		}
		return new GAIndividual(child);
	}

	public static GAIndividual xover1p(GAIndividual f,GAIndividual m){
	// 1-point cross over
		Random rng = new Random();
		int xpoint = 1 + rng.nextInt(f.genome_size - 1);
		float[] child = new float[f.genome_size];
		for(int i=0; i < xpoint; i++){
			child[i] = f.genome[i];
		}
		for(int i=xpoint; i < f.genome_size; i++){
			child[i] = m.genome[i];
		}
		return new GAIndividual(child);
	}

	public String toString(){

		String s = "[";
		for(int i=0; i < genome_size - 1; i++){
			s+= genome[i]+", ";
		}
		s+=genome[genome_size - 1] + "]";
		s+="  fitness = "+fitness;
		s+="  function value = "+ (1.0f / fitness - 1.0);

		return s;
	}

	private void evalFitness(){
	// greater fitness means better individual

		// De Jong's F1:
		// fitness = 1.0f / (1.0f + genome[0]*genome[0] + genome[1]*genome[1] + genome[2]*genome[2] );

		// De Jong's F2:
		// fitness = 1.0f / (1.0f + 100.0f*(genome[0]*genome[0]-genome[1])*(genome[0]*genome[0]-genome[1]) + (1.0f - genome[0])*(1.0f - genome[0]) );

//		Schaffer F6:
//		float tmp0 = genome[0]*genome[0] + genome[1]*genome[1];
//		float tmp1 = (float) Math.sin(Math.sqrt(tmp0));
//		float tmp2 = 1.0f + .001f*tmp0;
//		fitness = 1.0f / (1.0f + .5f + (tmp1*tmp1 -.5f) / (tmp2*tmp2));

// 		Ackey F1 (5 Dimentional):
//		float tmp0 = genome[0]*genome[0] + genome[1]*genome[1] + genome[2]*genome[2] + genome[3]*genome[3] + genome[4]*genome[4];
//		float tmp1 = (float)(Math.cos(2*Math.PI*genome[0]) + Math.cos(2*Math.PI*genome[1]) + Math.cos(2*Math.PI*genome[2]) + Math.cos(2*Math.PI*genome[3]) + Math.cos(2*Math.PI*genome[4]) );
//		fitness = 1.0f / (1.0f + 20.0f + (float)(Math.exp(1) - 20*Math.exp(-.2*Math.sqrt(tmp0/5.0f)) - Math.exp(tmp1/5.0f)));

//	  Rastrigin F1 (10 Dimentional)
		float sum = 100;
		for(int i=0; i < genome_size; i++)
			sum = sum + genome[i]*genome[i] - (float)(10.0*Math.cos(2*Math.PI*genome[i]));
		fitness = 1.0f / (1.0f + sum);
	}

  public static void main(String[] args) {
  }
}