coc.py

import numpy as np

low = 0
high = 9
chromes = 200
epoch = 2000
achivement = 0.58

cr = 0.4
mr = 0.3
corrector_rate = 0.5
master_fitness = ""
sample = np.random.randint(high, size=(10,2))

att = [430.0697618, 382.2041339, 315.9746825, 383.6665219, 456.2981481, 255.5777768, 269.8147513, 176.7710384, 127.984374, 84.70537173]
defe =[760, 712, 632, 560, 504, 528, 552, 406, 364, 270]



dna = []

def select(fs):
    p = np.random.uniform(0, 1)
    for i, f in enumerate(fs):
        if p <= 0:
            break
        p -= f
    return i

for ch in range(0,chromes):
    dna.append(np.random.randint(9, size=(10,2)))


for xxy in range(epoch):

    F_obj = []
    fitness = []

    fitness_total = 0
    p = []
    c = []

    new_gen = []
    cross = []

    for d in range(len(dna)):
        total = -5.6
        for dn in range(len(dna[d])):
            attacker = dna[d][dn][0]
            defence = dna[d][dn][1]

            v_attacker = att[attacker]
            v_defence = defe[defence]

            fight = v_attacker/v_defence
            # attack = achivement - fight
            # attack = np.power(attack,2)
            # attack = np.sqrt(attack)

            total += fight
            # print(attack)
        total = np.power(total,2)
        total = np.sqrt(total)
        F_obj.append(total)

    for d in range(len(F_obj)):
        fit = 1/(1+F_obj[d])
        fitness_total += fit
        fitness.append(fit)
        # print(fitness_total)

    # print(F_obj)

    for d in range(len(fitness)):
        prob = fitness[d]/fitness_total
        p.append(prob)

    # print('-----------')
    # print(p)


    for x in range(len(fitness)):
        sel = select(p)
        new_gen.append(dna[x])

    # print('-----------')

    dna = new_gen


    for x in range(len(dna)):
        r = np.random.uniform(0,1)

        mutation_cutoff = np.random.uniform(0,1)
        corrector = np.random.uniform(0,1)

        mpos = np.random.randint(9)
        mutation = np.random.randint(9)
        mp = np.random.randint(1)

        # if corrector < corrector_rate:
        #     for co in dna[x]:
        #         pos0 = dna[x][co][0][0]
        #         pos1 = dna[x][co][0][1]
        #         corrector_mutation = np.random.randint(-2, high=2)
        #
        #         if pos0-pos1 > 3:
        #             pos1 = pos0 + corrector_mutation
        #             dna[x][co][0][1] = pos1
        #         elif pos1 - pos0 > 3:
        #             pos0 = pos1 + corrector_mutation
        #             dna[x][co][0][0] = pos0


        if mutation_cutoff < mr:
            dna[x][mpos][mp] = mutation

        if r < cr:
            cross.append(x)


    for x in range(len(cross)):
        r = np.random.randint(len(cross))
        cpoint = np.random.randint(1,high=len(dna)-1)

        splice1 = dna[x][cpoint:]
        splice2 = dna[r][:cpoint]
        joint = np.concatenate((splice1,splice2), axis=0)

        dna[x] = joint

    maxfit = np.argmin(F_obj)
    master_fitness += str(F_obj[maxfit]) + "\n"
    if F_obj[maxfit] < 0.00001:
        break

minval = np.argmin(F_obj)

print(F_obj[minval])
print(minval)
print(dna[minval])
print(fitness[minval])

f = open('coc.csv','w')
f.write(master_fitness) # python will convert \n to os.linesep
f.close() # you can omit in most cases as the destructor will call it