我需要加速 python 中的嵌套循环，它会产生数万亿个循环

Question

filtradas is around 500000 elements and _13 around 2000000 the minimum time i have achieved is like 4 hours using cython, but I need to do it in less than an hour, how can i do it?

Both lists have string elements with 1, 2 or X the X can be changed for a 3

for i in filtradas:
        for x in _13:
            aciertos=0
            if i[0]==x[0]:
                aciertos+=1
            if i[1]==x[1]:
                aciertos+=1         
            if i[2]==x[2]:
                aciertos+=1
            if i[3]==x[3]:
                aciertos+=1
            if i[4]==x[4]:
                aciertos+=1
            if i[5]==x[5]:
                aciertos+=1         
            if i[6]==x[6]:
                aciertos+=1
            if i[7]==x[7]:
                aciertos+=1
            if i[8]==x[8]:
                aciertos+=1
            if i[9]==x[9]:
                aciertos+=1         
            if i[10]==x[10]:
                aciertos+=1
            if i[11]==x[11]:
                aciertos+=1
            if i[12]==x[12]:
                aciertos+=1
            if i[13]==x[13]:
                aciertos+=1
            if aciertos>=nroaciertos:
                filtradas13.append(i)
                break
    return filtradas13

Answer 1

numpy is worth a try. In this example I suspect that the time to convert to a numpy array isn't worth the compare but if your data was imported into numpy in the first place, it may go faster.

import numpy as np

_13_array = [np.array(list(x))for x in _13]
for i in filtradas:
    filtradas_array = np.array(list(i))
    for x in _13_array:
        aciertos = (x==filtradas_array).sum()
        if aciertos>=nroaciertos:
            filtradas13.append(i)
            break