random sampling

random_sample.py

#!/usr/bin/env python 

import sys 
import random
import argparse

parser = argparse.ArgumentParser(description='')
parser.add_argument('--ratio','-r', type=float, nargs='?'
                           ,required=True
                           ,help='ratio for random sampling')
parser.add_argument('--start_line', type=int, nargs='?',
                            default = 0,
                           help='start line')

args = parser.parse_args()
sys.stderr.write(str(args)+"\n")

def do(): 
    ratio = args.ratio
    start_line = args.start_line

    #sys.stderr.write("ratio: " + str(ratio) + "\n")
    #sys.stderr.write("start_line: " + str(start_line) + "\n")

    for i, li in enumerate(sys.stdin.readlines()):
        if i < start_line:
            sys.stdout.write(li)
        else:
            if ratio >= random.random():
                sys.stdout.write(li)


if __name__ == "__main__":
   do()