hits_by_trackable_users.py

#!/usr/bin/env python3
from __future__ import print_function

import ujson as json
#import demjson
import sys
from hashlib import md5
import datetime
from collections import defaultdict

def warning(*objs):
    print("WARNING: ", *objs, file=sys.stderr)
    
total_hits = 0   
ids = ['mainId', 'gadsid', 'hash_id', 'cxs', 'cxp']

trackable_hits = 0
tracker_ids = defaultdict(set)
tracker_hits = defaultdict(int)

while True:
  try:
    line = sys.stdin.readline()
    if not line:
       break
    jsn = json.loads(line)
    total_hits += 1
    for idname in ids:
      mainID = jsn[idname]
      if mainID in tracker_ids[idname]:
        tracker_hits[idname] += 1
      if mainID:
         tracker_ids[idname].add(mainID)
  except Exception as e:
    warning(e)
    
    
print("total hits: %s" % total_hits  )

for k,v in tracker_ids.items():
  hits = tracker_hits[k]
  print("%s\t perc of total: %0.2f hits: %s, unique: %s, ratio: %s" % (k, hits / total_hits, hits, len(v), hits / len(v)))