#!/usr/bin/python

# Minas Gjoka

import urllib,re, sys, time, random, datetime, time, signal, urllib2, cookielib, httplib, gzip, glob, os, csv
import Gnuplot, Gnuplot.funcutils
from scipy import *

try:
	name_apps_uid = sys.argv[1]
	name_users_appid = sys.argv[2]
except:
	print 'Usage:', sys.argv[0], 'apps_uid users_appid'
	sys.exit(1)

def calc_cdf(sorted_list):
	list_freq= {}
	for index in range(0,len(sorted_list)):
		list_freq[ sorted_list[index] ] =  list_freq.get( sorted_list[index] , 0) + 1
		
	elements = list_freq.keys()
	elements.sort( lambda x,y :cmp( x,y) )
	
	cdf = {}
	ccdf = {}
	n_entries = 0
	for index in range(0,len(elements)):
		n_entries = list_freq[  elements[index] ] + n_entries
		cdf[ elements[index] ] = float( n_entries)/len(sorted_list)
	set_cdf = [cdf[key] for key in elements]
	return [elements, set_cdf]


apps_uid = {}
users_appid = {}
nofapps_appid = {}    # count of the same applications
nofapps_frequency = {}   # frequency of number of applications per user


fapps_uid = open(name_apps_uid,"r")
for line in fapps_uid:
	fields = line.split()

	uid = fields[0]
	list = fields[1:len(fields)]
	
	apps_uid[uid] = list
fapps_uid.close()

print "Apps_uid loaded"

fusers_appid = open(name_users_appid,"r")
for line in fusers_appid:
	fields = line.split()
	
	appid = fields[0]
	list = fields[1:len(fields)]
	
	users_appid[appid] = list
fusers_appid.close()

print "Users_appid loaded"

for uid in apps_uid.keys():
	list = apps_uid.get( uid, [])
	nofapps_frequency[ len(list) ] = nofapps_frequency.get( len(list), 0) + 1

for appid in users_appid.keys():
	list = users_appid.get( appid, [])
	nofapps_appid[ appid ] =  len(list)

print "Calculate coverage"

# sort applications in decreasing order of popularity
list_appids = nofapps_appid.keys()
list_appids.sort(  lambda x,y :cmp( nofapps_appid[x] ,nofapps_appid[y]), reverse=True ) 

h = {}
count = 1
x_data = []
y_data = []
for appid in list_appids:
	listofuids = users_appid.get( appid, [])
	h.update( [(x,1) for x in listofuids])
	x_data.append( count )
	y_data.append( len(h.keys() ) )
	count = count + 1

print "Generating graphs"

#############Coverage#####################
fcoverage = open("coverage.dat","w")
n_allapps  = float(  len(users_appid.keys()) )
n_allusers = float(  len(apps_uid.keys()) )
print n_allapps, n_allusers
fcoverage.write("#RankedApps Coveredusers Perc_RankedApps(%d) Perc_Coveredusers(%d)\n" % (n_allapps, n_allusers) )
for i in range(len(x_data)):
	fcoverage.write("%d %d %f %f %s\n" % (x_data[i], y_data[i], x_data[i]/n_allapps, y_data[i]/n_allusers, list_appids[i]) )
fcoverage.close()


################CCDF of Total Installs#######################
installs_popularity = nofapps_appid.values()

installs_popularity.sort(reverse=False)
[elements, cdf_installs_popular] = calc_cdf(installs_popularity)

fout=open('ccdf-installs-datacrawled.dat',"w")
fout.write('')
for i in range(len(elements)):
	fout.write("%d %f\n" % (elements[i], 1-cdf_installs_popular[i]))
fout.close()


ranking = arange(1, len(installs_popularity)+1)
installs_popularity.sort(reverse=True)
fout = open("rank-installs-datacrawled.dat","w")
fout.write("#RankedApp TotalInstalls\n") 
for i in range(len(ranking)):
	fout.write('%d %d\n' % (ranking[i], installs_popularity[i] ) )
fout.close()

##################Histogram for number of apps####################
freq = nofapps_frequency.keys()
freq.sort(reverse=False)

fout = open("apps-peruser.dat","w")
fout.write("#Numofapps Instances\n")
for i in freq:
	fout.write("%d %d\n" % (i, nofapps_frequency[i]))
fout.close()