compare flow

compareflow.py

import sys

fname1 = sys.argv[1]
fname2 = sys.argv[2]

first_dict = {}
second_dict = {}

fp1 = open(fname1)
for line in fp1:
    items = line.strip().split("\t")
    path = items[0]
    flow = 0
    if items[1] != "NULL":
        flow = int(items[1])
    first_dict.update({path: flow})
fp1.close()

fp2 = open(fname2)
for line in fp2:
    items = line.strip().split("\t")
    path = items[0]
    flow = 0
    if items[1] != "NULL":
        flow = int(items[1])
    second_dict.update({path: flow})
fp2.close()

##check
first_keys = set(first_dict.keys())
second_keys = set(second_dict.keys())

shared_keys = first_keys & second_keys
first_only_keys = first_keys - second_keys
second_only_keys = second_keys - first_keys

for k in shared_keys:
    f_flow = first_dict[k]
    s_flow = second_dict[k]
    d_flow = s_flow - f_flow
    print("{0}\t{1}\t{2}\t{3}".format(k, f_flow, s_flow, d_flow))

for k in first_only_keys:
    f_flow = first_dict[k]
    print("{0}\t{1}\t{2}\t{3}".format(k, f_flow, 0, -f_flow))

for k in second_only_keys:
    s_flow = second_dict[k]
    print("{0}\t{1}\t{2}\t{3}".format(k, 0, s_flow, s_flow))

该程序的输入为hadoop汇总的结果，格式为：
path\tsum(length)