def dbconversation(tablename):
#creates edges from OP to mentioned
graph = nx.DiGraph()
#Plot following/follower network
#restrict to those with 3 tweets or more
mygetter = DBTweetGetter(None, None)
con = lite.connect("tweetsdb.db")
cur = con.cursor()
mintime = 1358090418
cur.execute("SELECT DISTINCT ScreenName FROM " + tablename +
" WHERE ConvertedTime>" + str(mintime) + " AND IsRetweet=0")
l = cur.fetchall()
users = []
for item in l:
users.append(item[0].lower())
cur.execute("SELECT Tweet, ScreenName FROM " + tablename +
" WHERE ConvertedTime>" + str(mintime) + " AND IsRetweet=0")
d = cur.fetchall()
lz = len(d)
z = 1
for item in d:
#print "Tweet " +str(z)+"/"+str(lz)
z += 1
if ("@" in item[0].lower()) and ("rt:" not in item[0].lower()):
#Continue until character not in valid set, then check if is user in set
#First count number of @s
names = []
c = item[0].lower().count("@")
start = 0
for i in range(c):
s = item[0].lower().index("@", start)
start = s + 1
k = 0
try:
j = item[0][start]
except:
print item[0]
while j in valid_characters:
k += 1
try:
j = item[0][start + k]
except:
j = "/"
names.append(item[0][start:start + k].lower())
for name in names:
if name.lower() in users and name.lower() != item[1].lower():
try:
graph[item[1].lower()][name.lower()]['weight'] += 1
except:
graph.add_edge(item[1].lower(), name.lower(), weight=1)
#graph.add_edge(item[1].lower(), name.lower())
print "Built graph"
nx.write_gml(graph, "newconv" + tablename + "nortdir.gml")
print "Wrote graph"
from classtweetgetter import DBTweetGetter
#from classtweetreader import DBTweetReader
import datetime
tags = [
"IPCC", "UNFCCC", "AR5", "WGII", "WGIII", "LTFchat", "Pages2k", "Pages",
"HadCRUT", "GISS"
]
for name in tags:
myTweets = DBTweetGetter("IPCCdb.db", name)
myTweets.query2("#" + name, 1600000, False)
#Write log file
# with open("log"+name+".txt", "a") as myfile:
# now = datetime.datetime.now()
# timestr=now.strftime("%d_%m_%H%M")
# myfile.write(timestr+"\n")
# readTweet=DBTweetReader("IPCCdb.db", tablename)
# print "Total number of tweets: " + str(readTweet.getNumberOfTweets(tablename))
# myfile.write("Total number of tweets: " + str(readTweet.getNumberOfTweets(tablename))+"\n")
# udict=readTweet.getUserDict(tablename)
# print "Total number of users: " + str(len(udict.keys()))
# myfile.write("Total number of users: " + str(len(udict.keys()))+"\n")
# for item in udict.items():
# if item[1]<3:
# del udict[item[0]]
# print "Total number of users with 3 tweets or more: " + str(len(udict.keys()))
# myfile.write("Total number of users with 3 tweets or more: " + str(len(udict.keys()))+"\n"
import sqlite3 as lite
import sys
from classtweetgetter import DBTweetGetter
mytweetgetter = DBTweetGetter(None, None)
con = lite.connect("tweetsdb.db")
cur = con.cursor()
ucon = lite.connect("userdb.db")
ucur = ucon.cursor()
tables = [
'htclimatechange', 'htclimate', 'htglobalwarming', 'ClimateChange',
'GlobalWarming'
]
names = []
for item in tables:
cur.execute("SELECT DISTINCT ScreenName FROM " + item)
a = cur.fetchall()
for item2 in a:
names.append(item2[0])
already = []
ucur.execute("SELECT ScreenName FROM usermap")
b = ucur.fetchall()
for item in b:
already.append(item[0])
deleted = []
i = 0
from classtweetgetter import DBTweetGetter
from classtweetreader import DBTweetReader
import datetime
tablename = "htclimate"
queryname = '#climate'
fname = "#climate"
myTweets = DBTweetGetter("tweetsdb.db", tablename)
myTweets.query(queryname, 160000, False)
#Write log file
with open("dblog" + fname + ".txt", "a") as myfile:
now = datetime.datetime.now()
timestr = now.strftime("%d_%m_%H%M")
myfile.write(timestr + "\n")
readTweet = DBTweetReader("tweetsdb.db", tablename)
print "Total number of tweets: " + str(
readTweet.getNumberOfTweets(tablename))
myfile.write("Total number of tweets: " +
str(readTweet.getNumberOfTweets(tablename)) + "\n")
udict = readTweet.getUserDict(tablename)
print "Total number of users: " + str(len(udict.keys()))
myfile.write("Total number of users: " + str(len(udict.keys())) + "\n")
for item in udict.items():
if item[1] < 3:
del udict[item[0]]
print "Total number of users with 3 tweets or more: " + str(
len(udict.keys()))
myfile.write("Total number of users with 3 tweets or more: " +
str(len(udict.keys())) + "\n")
import sqlite3 as lite
import sys
import numpy as np
from classtweetgetter import DBTweetGetter
from time import sleep
mygetter=DBTweetGetter(None,None)
con=lite.connect("tweetsdb.db")
cur=con.cursor()
ucon=lite.connect("userdb.db")
ucur=ucon.cursor()
crawlers=[]
chainlengths=[]
nsame=[]
noded={}
chainlfile=open("chaindata.txt","w")
nsamefile=open("nsamedata.txt","w")
class chainCrawler(object):
#note this method will repeat chains if there is V structure, should be minimal effect
def __init__(self, node, n):
self.n=n
self.node=node
self.stopwalk=False
def walk(self):
while self.stopwalk==False:
self.step()
return 0
def step(self):
def dbplotffnetwork():
graph = nx.DiGraph()
#Plot following/follower network
#restrict to those with 3 tweets or more
mintime = 1358090418
maxtime = 1363963163
mygetter = DBTweetGetter(None, None)
con = lite.connect("tweetsdb.db")
cur = con.cursor()
ucon = lite.connect("userdb.db")
ucur = ucon.cursor()
tusers = []
users = []
cur.execute(
"SELECT ScreenName FROM htglobalwarming WHERE ConvertedTime > " +
str(mintime) + " AND ConvertedTime < " + str(maxtime) +
" COLLATE NOCASE")
temp = cur.fetchall()
for item in temp:
tusers.append(item[0].lower())
for item in tusers:
if not (item in users):
if tusers.count(item) > 7:
users.append(item)
print len(users)
# cur.execute("SELECT ScreenName FROM htclimatechange WHERE ConvertedTime > "+str(mintime)+" AND ConvertedTime < " + str(maxtime) + " COLLATE NOCASE")
# temp=cur.fetchall()
# tusers=[]
# for item in temp:
# tusers.append(item[0].lower())
# for item in tusers:
# if not (item in users):
# if tusers.count(item)>29:
# users.append(item)
# print len(users)
# cur.execute("SELECT ScreenName FROM htagw WHERE ConvertedTime > "+str(mintime) + " COLLATE NOCASE")
# temp=cur.fetchall()
# tusers=[]
# for item in temp:
# tusers.append(item[0].lower())
# for item in tusers:
# if not (item in users):
# if tusers.count(item)>2:
# users.append(item)
# print len(users)
#aim for 380
#sys.exit("Hammertime")
i = 0
try:
users.remove("undercoverzen")
users.remove("jivelad")
users.remove("anabananazavala")
#TODO Formalise this
except:
pass
for user in users:
print "User " + str(i) + "/" + str(len(users))
i += 1
#For each user check which other users are in friends, followers
ucur.execute("SELECT FriendId FROM friends WHERE ScreenName='" +
user.lower() + "' COLLATE NOCASE")
frl = []
temp = ucur.fetchall()
skip = False
if len(temp) == 0:
#get friends
print "Downloading friends for " + user.lower()
friendslist = mygetter.getFriends(user.lower(), [], -1)
if friendslist != "FAIL":
for friend in friendslist:
ucur.execute("INSERT INTO friends VALUES('" +
user.lower() + "'," + str(friend) + ")")
frl = friendslist
else:
skip = True
try:
users.remove(user.lower())
except:
pass
sleep(10)
else:
for item in temp:
frl.append(item[0])
ucur.execute("SELECT FollowerId FROM followers WHERE ScreenName='" +
user + "' COLLATE NOCASE")
fol = []
temp = ucur.fetchall()
skip = False
if len(temp) == 0:
#get friends
print "Downloading followers for " + user.lower()
followerslist = mygetter.getFollowers(user.lower(), [], -1)
if followerslist != "FAIL":
for follower in followerslist:
ucur.execute("INSERT INTO followers VALUES('" +
user.lower() + "'," + str(follower) + ")")
fol = followerslist
else:
skip = True
try:
users.remove(user.lower())
except:
pass
sleep(10)
else:
for item in temp:
fol.append(item[0])
ucon.commit()
if skip == False:
graph.add_node(user.lower())
for other in users:
skip2 = False
ucur.execute("SELECT UserId FROM usermap WHERE ScreenName='" +
other.lower() + "' COLLATE NOCASE")
temp = ucur.fetchall()
if len(temp) == 0:
#get ID from web
print "Downloading userid for " + other.lower()
x = mygetter.getIDfromUser(other.lower())
if x != "FAIL":
ucur.execute("INSERT INTO usermap VALUES('" +
other.lower() + "'," + x + ")")
sid = x
ucon.commit()
else:
try:
users.remove(other.lower())
except:
pass
skip2 = True
sleep(10)
else:
sid = temp[0][0]
if skip2 == False:
if sid in fol:
graph.add_edge(other.lower(), user.lower())
if sid in frl:
graph.add_edge(user.lower(), other.lower())
print "Built graph"
nx.write_gml(graph, "newfriendfollowerhtccgt29.gml")
ucon.commit()
con.close()
ucon.close()
print "Wrote graph"
from classtweetgetter import DBTweetGetter
from classtweetreader import DBTweetReader
import datetime
myTweets = DBTweetGetter("tweetsdb.db", "htclimatechange")
myTweets.query('#climatechange', 160000, False)
#Write log file
with open("dblog#climatechange.txt", "a") as myfile:
now = datetime.datetime.now()
timestr = now.strftime("%d_%m_%H%M")
myfile.write(timestr + "\n")
readTweet = DBTweetReader("tweetsdb.db", "htclimatechange")
print "Total number of tweets: " + str(
readTweet.getNumberOfTweets("htclimatechange"))
myfile.write("Total number of tweets: " +
str(readTweet.getNumberOfTweets("htclimatechange")) + "\n")
udict = readTweet.getUserDict("htclimatechange")
print "Total number of users: " + str(len(udict.keys()))
myfile.write("Total number of users: " + str(len(udict.keys())) + "\n")
for item in udict.items():
if item[1] < 3:
del udict[item[0]]
print "Total number of users with 3 tweets or more: " + str(
len(udict.keys()))
myfile.write("Total number of users with 3 tweets or more: " +
str(len(udict.keys())) + "\n")
#modify this script
# with open("gettweetsClimateChange.py", "r") as myfile:
# mytext=myfile.read()
