Large problems need many iterations. The current implementation keeps a solution for every iteration in memory which leads to problems for large problems.

Only store best and current candidate.

There is a Rastrigin example on this package README file. The example output for that example is reported on the README file as:

Try the same example

I run the same example on the README file, three times. Each time I get a different result.

1st try

finished Simulated Annealing in: 227.100175ms
finished Simulated Annealing, Final Result/State: [0.0029467658259519063 -0.0019141863449331415
finished Simulated Annealing, Final Value/Energy: 0.0024495962011314987

2nd try

finished Simulated Annealing in: 240.450095ms
finished Simulated Annealing, Final Result/State: [-0.0005143562820309873 -0.0024060189297087885
finished Simulated Annealing, Final Value/Energy: 0.0012009426091914577

3rd try

finished Simulated Annealing in: 190.875726ms
finished Simulated Annealing, Final Result/State: [0.0009942218440814055 0.0001947203504867579
finished Simulated Annealing, Final Value/Energy: 0.00020362763283543472

Question

Is it expected to get different results? Is such a difference acceptable? Am I missing something? Thanks.

GA and ACO use the weightedChoice method. Currently it uses a linear search strategy with O(N) performance. However a binary search with O(log(N)) could be used whenever multiple selections from the same weights are made.

I found a python implementation here

def prepare_binary_search(weights):
    # Computing the running totals takes O(N) time
    running_totals = list(itertools.accumulate(weights))
        
    def weighted_random():
        target_distance = random()*running_totals[-1]
        low, high = 0, len(weights)
        while low < high:
            mid = int((low + high) / 2)
            distance = running_totals[mid]
            if distance < target_distance:
                low = mid + 1
            elif distance > target_distance:
                high = mid
            else:
                return mid
        return low

    return weighted_random