In [1]:

    

import numpy as np


    

In [7]:

    

np.random.seed(50)


    

In [8]:

    

indexlist = range(0,40428967)


    

In [9]:

    

np.random.shuffle(indexlist)


    

In [10]:

    

indexset = set(indexlist[:20214483])


    

In [11]:

    

%%time
i = 0
with open('train1.txt', 'wb') as outfile:
    for line in open("train.txt"):
        if i in indexset:
            outfile.write(line)
        i+=1


    

    




CPU times: user 33 s, sys: 10.9 s, total: 43.9 s
Wall time: 3min 40s

In [25]:

    

%%time
i = 0
with open('train1.fm', 'wb') as outfile:
    for line in open("train.fm"):
        if i in indexset:
            outfile.write(line)
        i+=1


    

    




CPU times: user 27.6 s, sys: 5.2 s, total: 32.8 s
Wall time: 2min 14s

In [ ]: