## subsampling from extended.bed file
## need to pass a parameter

import pandas as pd
import random
import numpy as np
import sys

file=sys.argv[1]
outname=file[ :-13]

sample = pd.read_csv('%s' %file, sep='\t', header=None)

n = len(sample)
point = (5, 10, 20, 30, 40, 50, 60, 70, 80, 90)

for p in point:
   size = int( n * p / 100)             #total reads to take
   take = random.sample(range(0,n), size)        
   take.sort( )                   
   content= sample.loc[take, ]          # subset of sample 
   np.savetxt('%s_sample%d.extended.bed' %(outname, p), content, fmt='%s', delimiter='\t')