make test dataset

make_data.py

#!/usr/bin/env python

import os

begin = 1
end = 3000000
inc = 3000000
for j in xrange(3):
    with open('%s.txt' % (j + 1), 'w') as wf:
        for i in xrange(begin, end + 1):
            print >> wf, i

    begin += inc
    end += inc


os.system("tar czvf all.tgz 1.txt 2.txt 3.txt")