data_nasa.py

gistfile1.txt

import sys
import getopt
import csv
import argparse

def main(argv):
    args = getopt.getopt(argv, "")
    args = str(args[1][0])
    print args
    def get_cnt(lVals):
        d = dict(zip(lVals, [0] * len(lVals)))
        for x in lVals:
            d[x] += 1
        return d

    print "Im inno"
    url = []
    with open(args) as tsvfile:
    # with open("sample_100.tsv") as tsvfile:
        tsvreader = csv.reader(tsvfile, delimiter="\t")
        for line in tsvreader:
            url.append(line[4:5][0])
    url_map = get_cnt(url)
    print get_cnt(url)
    writer = csv.writer(open('dict.csv', 'wb'))
    for key, value in url_map.items():
        writer.writerow([key, value])

if __name__ == "__main__":
    main(sys.argv[1:])