array-genetics.js

/*
//Sample genetic-ish algorithm
Original:
[[0, 0
 [0 .
     . 
      . ]
       0]]

Desired:
[[x1,
 [   x2
       ...  ]
          x8]]
such that x1 + x2 + ... + x8 >= 10000
(Yes I know, 8 variables means 7 degrees of freedom, but this algorithm discounts that
and instead makes a nice pretty distribution)
*/

var arr = [0, 0, 0, 0, 0, 0, 0, 0];
for(var i = 0; i < arr.length; i++) {
	arr[i] = [0, 0, 0, 0, 0, 0, 0, 0];
}

var generation = 0, instability = 2, population = 600, targetFitness = 10000;

function getFitness(tempArr) {
	var fitness = 0;
	for(var i = 0; i < tempArr.length; i++) {
		fitness += tempArr[i][i];
	}
	return fitness;
}

function mutate(tempArr) {
	for(var i = 0; i < tempArr.length; i++)
		for(var j = 0; j < tempArr[i].length; j++)
			tempArr[i][j] += Math.round(Math.random() * instability - instability/2);
}

function copyArr(tempArr) {
	var newArr = tempArr.slice(0);
	for(var i = 0; i < tempArr.length; i++) { //
		newArr[i] = tempArr[i].slice(0);
	}
	return newArr;
}

var currFitness = 0;
while(currFitness < targetFitness) {
	var bestFitness = 0, bestArr = arr, fit, tempArr;
	for(var p = 0; p < population; p++) {
		tempArr = copyArr(arr); //Duplicate array
		mutate(tempArr);
		fit = getFitness(tempArr);
		if(fit > bestFitness) {
			bestFitness = fit;
			bestArr = tempArr;
		}
	}
	arr = bestArr;
	currFitness = bestFitness;
	generation++;
	if(generation % 1000 === 0)
		console.log(generation + ": " + currFitness);
}
console.log(generation + ": " + currFitness);
console.log(arr);