def getPyhmeter(tweet_text):
#TODO: remove URL and other stuff to clean text
pattern = re.compile('(@\w*)|(\|.*)|(#\s*)')
removedAt = pattern.sub('', tweet_text)
tk = TweetTokenizer()
tokens = tk.tokenize(removedAt)
return pyhmeter.HMeter(tokens, file, 1)
def main():
def printTweet(descr, t):
print(descr)
print("Username: %s" % t.username)
print("Retweets: %d" % t.retweets)
print("Text: %s" % t.text)
print("Mentions: %s" % t.mentions)
print("Hashtags: %s\n" % t.hashtags)
# data=pd.read_csv(os.getcwd()+"/followers.csv")
# # print(data)
# l=[]
# print(data["screen_name"][0][35:(len(data["screen_name"][0])-(len(data["screen_name"][0])-36)/2)-4])
# # print(data["screen_name"][1])
# f=open(os.getcwd()+"/followers.txt","a")
# for i in range(len(data["screen_name"])):
# f.write(data["screen_name"][i][35:(len(data["screen_name"][i])-(len(data["screen_name"][i])-36)/2)-4]+"\n")
# l.append(data["screen_name"][i][35:(len(data["screen_name"][i])-(len(data["screen_name"][i])-36)/2)-4])
# # print(l)
data = pd.read_csv(os.getcwd() + "/followers.txt", header=None)
# print(data[0])
sample_scores = pyhmeter.load_scores("Data_Set_S1.txt")
scores = []
dates = []
days = ["%.2d" % i for i in range(1, 30)]
cnt = 0
for month in ["%.2d" % i for i in range(1, 13)]:
for day in range(len(days) - 1):
date = "2019-" + str(month) + "-" + str(days[day])
next_date = "2019-" + str(month) + "-" + str(days[day + 1])
day_score = 0
for user in data[0][:100]:
print("querying...")
cnt += 1
print(cnt)
tweetCriteria = got.manager.TweetCriteria().setUsername(
user).setQuerySearch("ClimateChange").setSince(
date).setUntil(next_date).setMaxTweets(1)
sum = 0
for i in got.manager.TweetManager.getTweets(tweetCriteria):
# printTweet("### Example 1 - Get tweets by username [barackobama]", i)
h = pyhmeter.HMeter(list((i.text).split()), sample_scores)
sum += h.happiness_score()
# dates.append(i.date)
# print(i.date)
# l.append(h.happiness_score())
# print(h.happiness_score())
# cnt+=1
# print(cnt)
sum /= i
day_score += sum
day_score /= 100
dates.append(date)
scores.append(day_score)
datetimes = matplotlib.dates.date2num(dates)
plt.plot_date(datetimes, scores, marker=None, linestyle="-")
plt.xticks(rotation=70)
plt.xlabel("time")
plt.ylabel("hedonometer score")
plt.savefig(
"/home/saksham/Twitter_Project/GetOldTweets-python-master/test3.png")
plt.show()
def main():
def printTweet(descr, t):
print(descr)
print("Username: %s" % t.username)
print("Retweets: %d" % t.retweets)
print("Text: %s" % t.text)
print("Mentions: %s" % t.mentions)
print("Hashtags: %s\n" % t.hashtags)
data = pd.read_csv(os.getcwd() + "/followers.txt", header=None)
# print(data[0])
sample_scores = pyhmeter.load_scores("Data_Set_S1.txt")
scores = []
dates = []
cnt = 0
average_score = 0
for month in ["%.2d" % i for i in range(1, 13)]:
monthly_score = 0
user_cnt = 0
cnt = 0
for user in data[0].sample(n=2000):
tweetCriteria = got.manager.TweetCriteria().setUsername(
user).setQuerySearch("ClimateChange").setSince(
"2019-" + month + "-01").setUntil("2019-" + month +
"-30").setMaxTweets(10)
# sum=0
# cnt=0
for i in got.manager.TweetManager.getTweets(tweetCriteria):
# printTweet("### Example 1 - Get tweets by username [barackobama]", i)
user_cnt += 1
h = pyhmeter.HMeter(list((i.text).split()), sample_scores)
if h.happiness_score() is not None:
monthly_score += h.happiness_score()
print(cnt)
cnt += 1
print(user)
print(h.happiness_score())
# dates.append(i.date)
# print(i.date)
# l.append(h.happiness_score())
# print(h.happiness_score())
if cnt != 0:
# monthly_score+=sum
monthly_score /= cnt
print(monthly_score)
scores.append(monthly_score)
average_score += monthly_score
dates.append(month)
# datetimes=matplotlib.dates.date2num(dates)
# plt.plot_date(datetimes,scores,marker=None,linestyle="-")
average_score /= len(scores)
print(scores, dates)
plt.plot(dates, scores, label="actual data", color="red")
plt.axhline(y=average_score, label="average score")
# plt.plot(average_score)
plt.xticks(rotation=70)
plt.xlabel("time(month)")
plt.ylabel("hedonometer score")
plt.legend()
# plt.title("")
plt.savefig(
"/home/saksham/Twitter_Project/GetOldTweets-python-master/test3.png")
plt.show()
dates = []
tweets = []
url_rem = []
neutral_rem = []
normal = []
url_score = []
neutral_score = []
# tweetCriteria=got.manager.TweetCriteria().setUsername("barackobama").setQuerySearch("ClimateChange").setSince('2019-03-01').setUntil('2019-12-30').setMaxTweets(10000)
tweetCriteria = got.manager.TweetCriteria().setUsername(
"KremlinRussia_E").setQuerySearch("Climate").setSince(
"2009-12-30").setUntil("2019-12-30").setMaxTweets(100000)
print(got.manager.TweetManager.getTweets(tweetCriteria))
for i in got.manager.TweetManager.getTweets(tweetCriteria):
text = re.sub(r'[^\x00-\x7F]+', ' ', i.text)
tweets.append(text)
h = pyhmeter.HMeter(list(i.text.split()), sample_scores)
normal.append(h.happiness_score())
h.deltah = 1.0
neutral_score.append(h.happiness_score())
neutral_rem.append(h.matchlist)
dates.append(i.date)
data = pd.DataFrame(
list(zip(tweets, neutral_rem, normal, neutral_score)),
columns=['normal', 'cleaned_tweets', 'no_clean', 'cleaned_score'])
data.to_csv(os.getcwd() + "/putin(2009-2020)cleaned(deltah=1.0).csv", sep="|")
datetimes = matplotlib.dates.date2num(dates)
plt.plot_date(datetimes, neutral_score, marker=None, linestyle="-")
plt.xticks(rotation=70)
plt.xlabel("time")
plt.ylabel("hedonometer score")