timeInterval = ["", ""]
fashion = []
awards = []
with con:
cur = con.cursor()
for q in range(len(sql_get_begEnd)):
cur.execute(sql_get_begEnd[q])
row = cur.fetchone()
timeInterval[q] = time.strptime(
re.search(r'\s([\w:]+)',
str(row[0])).group(1).encode('ascii', 'ignore'),
"%H:%M:%S")
eventTimeline = timeline(timeInterval[0], timeInterval[1])
# commence querying for award cats/winners
for query in sql_awards:
cur.execute(query)
rows2 = cur.fetchall()
for row in rows2:
# RT from @eonline or @goldenglobes
result1 = re.search(
r'RT @\w+: ([Bb]est [\w\s,]*)[:-] (Drama -|Comedy or Musical -|Musical or Comedy)?([A-Za-z\s\"-]*)(http)?',
row[1])
if result1:
# get timestamp
tStamp = time.strptime(
re.search(r'\s([\w:]+)',
from textblob import TextBlob
import re
import tweepy
import csv
from access_grant import *
from polarity import *
from timeline import *
from search import *
access, api = api_1()
print(access)
print("1.Enter 1 if you want to analyze the Timeline of any User")
print(
"2.Enter 2 if you want analyze Public Actions on Twitter towards the User")
options = int(input('Options:'))
if options == 1:
timeline = timeline()
elif options == 2:
search()
else:
print("Wrong Option")
else:
print("Usage: " + sys.argv[0] + " percentdone <item_num> [PERCENT]")
elif (action == "due"):
if (len(args) > 1 and args[0].isdigit()):
rmExtension(int(args[0]), "due")
addExtension(int(args[0]), "due", formatDate(" ".join(args[1:])))
elif (len(args) > 0 and args[0].isdigit()):
rmExtension(int(args[0]), "due")
else:
print("Usage: " + sys.argv[0] + " due <item_num> [DATE]")
elif (action in ["starts","start"]):
if (len(args) > 1 and args[0].isdigit()):
rmExtension(int(args[0]), "start")
addExtension(int(args[0]), "start", formatDate(" ".join(args[1:])))
elif (len(args) > 0 and args[0].isdigit()):
rmExtension(int(args[0]), "start")
else:
print("Usage: " + sys.argv[0] + " starts <item_num> [DATE]")
elif (action == "wait"):
if (len(args) > 1 and args[0].isdigit()):
rmExtension(int(args[0]), "wait")
addExtension(int(args[0]), "wait", " ".join(args[1:]))
elif (len(args) > 0 and args[0].isdigit()):
rmExtension(int(args[0]), "wait")
else:
print("Usage: " + sys.argv[0] + " wait <item_num> [for what]")
elif (action == "timeline" or action == "t"):
timeline()
else:
help()
con = lite.connect("gg_tweets.sqlite3")
timeInterval = ["",""]
fashion = []
awards = []
with con:
cur = con.cursor()
for q in range(len(sql_get_begEnd)):
cur.execute(sql_get_begEnd[q])
row = cur.fetchone()
timeInterval[q] = time.strptime(re.search(r'\s([\w:]+)', str(row[0])).group(1).encode('ascii', 'ignore'), "%H:%M:%S")
eventTimeline = timeline(timeInterval[0],timeInterval[1])
# commence querying for award cats/winners
for query in sql_awards:
cur.execute(query)
rows2 = cur.fetchall()
for row in rows2:
# RT from @eonline or @goldenglobes
result1 = re.search(r'RT @\w+: ([Bb]est [\w\s,]*)[:-] (Drama -|Comedy or Musical -|Musical or Comedy)?([A-Za-z\s\"-]*)(http)?', row[1])
if result1:
# get timestamp
tStamp = time.strptime(re.search(r'\s([\w:]+)', str(row[0])).group(1).encode('ascii', 'ignore'), "%H:%M:%S")
# get raw data
rawCat = result1.group(1).encode('ascii', 'ignore').strip("- ")
else:
print "Usage: " + sys.argv[0] + " percentdone <item_num> [PERCENT]"
elif (action == "due"):
if (len(args) > 1 and args[0].isdigit()):
rmExtension(int(args[0]), "due")
addExtension(int(args[0]), "due", formatDate(" ".join(args[1:])))
elif (len(args) > 0 and args[0].isdigit()):
rmExtension(int(args[0]), "due")
else:
print "Usage: " + sys.argv[0] + " due <item_num> [DATE]"
elif (action in ["starts","start"]):
if (len(args) > 1 and args[0].isdigit()):
rmExtension(int(args[0]), "start")
addExtension(int(args[0]), "start", formatDate(" ".join(args[1:])))
elif (len(args) > 0 and args[0].isdigit()):
rmExtension(int(args[0]), "start")
else:
print "Usage: " + sys.argv[0] + " starts <item_num> [DATE]"
elif (action == "wait"):
if (len(args) > 1 and args[0].isdigit()):
rmExtension(int(args[0]), "wait")
addExtension(int(args[0]), "wait", " ".join(args[1:]))
elif (len(args) > 0 and args[0].isdigit()):
rmExtension(int(args[0]), "wait")
else:
print "Usage: " + sys.argv[0] + " wait <item_num> [for what]"
elif (action == "timeline" or action == "t"):
timeline()
else:
help()
from timeline import *
from sklearn.cluster import KMeans
from sklearn.cluster import AffinityPropagation as AP
from sklearn.preprocessing import MinMaxScaler
import numpy as np
import matplotlib.pyplot as plt
api = login()
T, tweets = timeline(api) #buscamos datos del timeline
T = cleanse_data(
T
) #limpiamos los datos, es decir, preparar la data para normalizacion. En este caso asumimos que los booleanos se expresan como -1 (None), 0 (False) y 1 (True)
T = np.array(
[T[0], T[-1]]
) #en este caso podemos utilizar dos rasgos, utilizaremos el numero de retweets y si el tweet fue retwitteado o no ya que estos valores son numericos
T[0] = MinMaxScaler().fit_transform(
np.float64(T[0])
) #como los rasgos de retwitteado o no ya fueron normalizados, procedemos a normalizar la cantidad de retweets para que tenga valores oscilando entre 0 y 1
T = make_2d_mat(T) #tenemos que convertir los vectores a matrices
def kmeans(T, metodo):
model = KMeans(
n_clusters=2, init=metodo, precompute_distances=True
) #el numero de clusters tiene que ser mayor o igual al numero de rasgos que utilizamos, inicializamos con el metodo elegido (k-means++ y random) y establecemos que el algoritmo debe computar las distancias
model.fit(
