next_ind_order = individual_order
# エリート戦略
elite = fitness.index(max(fitness))
# 世代ごとの一番高い適応度とエリート番地
print(str(i + 1) + "世代目：" + str(maxinds_fit[i]) + "\tエリート番地：" + str(elite))
f.write(str(i + 1) + "," + str(max(fitness)) + "\n")
next_ind_path[0] = individual_path[elite]
next_ind_order[0] = individual_order[elite]
for j in range(1 , 100):
sel_num = rl.selection(fitness)
newinds = []
newinds = co.crossover(individual_order[sel_num[0]] , individual_order[sel_num[1]])
next_ind_order[j] = newinds
next_ind_path[j] = cv.convertOTP(next_ind_order[j])
individual_order = next_ind_order
individual_path = next_ind_path
print(np.array(individual_path).reshape(inds , len_dat))
print()
f.close()
#kame.path(individual_path , dat , fitness , len_dat)
next_path = np.array([0 for col in range(inds) for row in range(len_dat)]).reshape(inds, len_dat)
next_order = np.array([0 for col in range(inds) for row in range(len_dat)]).reshape(inds, len_dat)
# 順番に選択
for j in range(inds):
j_path = individual_path[j]
j_order = individual_order[j]
# ランダム3つ
select_path, select_order = sri.select(inds, individual_path, individual_order, j, len_dat)
# 差分変異親
mutant_order = np.array(cm.mutant(select_order, S_factor))
# 交叉して子(Trial)の生成
trial_order = co.crossover(j_order, mutant_order)
trial_path = cv.convertOTP(trial_order)
# 親と子で競争
fitness[j], next_path[j], next_order[j] = ev.evaluation(dat, j_path, j_order, trial_path, trial_order)
individual_path = next_path
individual_order = next_order
print(str(i + 1) + "世代目：" + str(max(fitness)) + "\n" + str(individual_path[fitness.index(max(fitness))]))
f.write(str(i + 1) + "," + str(max(fitness)) + "\n")
f.close()
print(individual_path[fitness.index(max(fitness))])
