def sent(comment, status):
analyzer = SentimentIntensityAnalyzer()
if (comment == ''):
return 'sample_value'
#initializations
flag = ''
val = 0
#values range from -1 to 1
p_threshold = 0.350
n_threshold = -0.150
result = analyzer.polarity_scores(comment)
val = result['compound']
print("result= ")
print(result)
if (val >= p_threshold):
flag = 'Green'
elif (val <= n_threshold):
flag = 'Red'
elif (val < p_threshold and val > n_threshold):
flag = 'Amber'
if (flag == status):
return 'done'
else:
return 'close'
def getInterestingNewsInRange(keyword, startdate, enddate):
db = mongodb_client.get_db()
startdate = dateutil.parser.parse(startdate)
enddate = dateutil.parser.parse(enddate)
interesting_news = db[NEWS_TABLE_NAME].find({
'publishedAt':{'$gte' : startdate, '$lte':enddate },
'$or' : [ { "title" : {'$in': [re.compile(keyword, re.IGNORECASE)]} }, { "description" : {'$in': [re.compile(keyword, re.IGNORECASE)]} } ]
})
interesting_news = list(interesting_news)
filtered_news = []
for news in interesting_news:
sentence_list = tokenize.sent_tokenize(news['text'])
selected_sentence_list = []
for sentence in sentence_list:
if (re.compile(r'\b({0})\b'.format(keyword), flags=re.IGNORECASE).search(sentence) != None):
selected_sentence_list.append(sentence)
analyzer = SentimentIntensityAnalyzer()
paragraphSentiments=0.0
if len(selected_sentence_list) != 0:
for sentence in selected_sentence_list:
vs = analyzer.polarity_scores(sentence)
paragraphSentiments += vs["compound"]
news['rate'] = str(round(paragraphSentiments/len(selected_sentence_list), 4))
filtered_news.append(news)
return json.loads(dumps(filtered_news))
def analyze_sentiment(text):
analyzer = SentimentIntensityAnalyzer()
sentiment = analyzer.polarity_scores(text)
if (sentiment["compound"] > .005) or (sentiment["pos"] > abs(
sentiment["neg"])):
return "Bullish"
elif (sentiment["compound"] < -.005) or (abs(sentiment["neg"]) >
sentiment["pos"]):
return "Bearish"
else:
return "Neutral"
# def analyze_text(item):
# # BASE_URL = "http://localhost:3000/comments"
# isPost = type(item) == praw.models.reddit.submission.Submission
# isDict = type(item) == dict
# # awards = ''
# if(isDict):
# text = item['body']
# id = item['id']
# time = item['created_utc']
# score = item['score']
# parent = item['parent_id']
# if parent.startswith("t"):
# parent = parent[3:]
# else:
# text = item.body
# time = item.created_utc
# id = item.id
# score = item.score
# parent = item.link_id[3:]
# # awards = item.all_awardings
# for word in text.split():
# word = word.strip(punctuation)
# # Tickers of len<2 do not exist
# if (len(word) < 2):
# continue
# # Does word fit the ticker criteria
# if word.isupper() and len(word) != 1 and (word.upper() not in gl.COMMON_WORDS) and len(word) <= 5 and word.isalpha() and (word.upper() in gl.TICKERS):
# # Checks to see if the ticker has been cached.
# # url = "http://localhost:3000/id/" + id
# r = requests.get(url= BASE_URL + "/id/" + id)
# if(r.status_code == 200):
# continue
# sentiment = analyze_sentiment(text)
# # print(score)
# data = {
# "comment_id" : id,
# "comment_date" : time,
# "ticker" : word,
# "parent_post" : parent,
# "body" : text,
# "score" : score,
# "sentiment" : sentiment
# }
# r = requests.post(url = BASE_URL+"/comments", data = data)
def analyze(body):
analyzer = SentimentIntensityAnalyzer()
sentiment = analyzer.polarity_scores(body)
if (sentiment["compound"] > .005) or (sentiment["pos"] > abs(
sentiment["neg"])):
return "Bullish"
elif (sentiment["compound"] < -.005) or (abs(sentiment["neg"]) >
sentiment["pos"]):
return "Bearish"
else:
return "Neutral"
def analyze_sentiment(self):
analyzer = SentimentIntensityAnalyzer()
neutral_count = 0
for text in self.bodies:
sentiment = analyzer.polarity_scores(text)
if (sentiment["compound"] > .005) or (sentiment["pos"] > abs(sentiment["neg"])):
self.pos_count += 1
elif (sentiment["compound"] < -.005) or (abs(sentiment["neg"]) > sentiment["pos"]):
self.neg_count += 1
else:
neutral_count += 1
self.bullish = int(self.pos_count / len(self.bodies) * 100)
self.bearish = int(self.neg_count / len(self.bodies) * 100)
self.neutral = int(neutral_count / len(self.bodies) * 100)
def evaluate_reaction(identifiers, corpus):
relevant_twts = [
twt_dict['tweet_text']
for twt_dict in filter_tweets(identifiers, corpus)
]
num_twts = len(relevant_twts)
metrics = numpy.ndarray(shape=(num_twts, 4))
sid = SentimentIntensityAnalyzer()
for i, twt in enumerate(relevant_twts):
ss = sid.polarity_scores(twt)
metrics[i, :] = [ss['compound'], ss['pos'], ss['neg'], ss['neu']]
score_order = ['compound', 'positive', 'negative', 'neutral']
avg_scores = numpy.average(metrics, axis=0)
for i in range(0, 4):
print score_order[i] + ": " + str(avg_scores[i])
return metrics
def insert_into_mysql(list_of_dict):
cur, conn = create_conn_to_mysql()
analyzer = SentimentIntensityAnalyzer()
for i in list_of_dict:
a = ''
b = ''
xyz = i['name']
try:
a, b = xyz.split(',')
except:
a = xyz
z = "\"" + i['comment'] + "\""
s = date_error(i['date'])
w = i['address']
print(">>>>> returned date", s)
#print(tpl[1])
vs = analyzer.polarity_scores(i['comment'])
#print(str(vs))
sen = str(vs)
tpl = [xyz, z, s, sen]
insert_into_table(cur, tpl)
close_connection(cur, conn)
dict = {}
dict['title'] = item.find("h3").find(text=True).strip('\t\n')
dict['date'] = item.find("p", {
"class": "review-author"
}).find("span", {
"class": "author-wrapper"
}).find(text=True).strip('\t\n')
dict['author'] = item.find("p", {
"class": "review-author"
}).find("span", {
"itemprop": "name"
}).find(text=True).strip('\t\n')
dict['comment'] = item.find("p", {
"class": "review-description"
}).find(text=True).strip('\t\n')
vs = analyzer.polarity_scores(dict['comment'])
dict['sentiment'] = str(vs)
xyz.append(dict)
#list_dict = xyz
#print(xyz)
con = pymysql.connect(user='******',
password='******',
database='customerreviews',
charset='utf8')
cursor = con.cursor()
DB_NAME = 'customerreviews'
for i in xyz:
w = i['author']
v = i['title']
x = i['date']
from vaderSentiment import SentimentIntensityAnalyzer
#sentences = ["now i am not feeling good"]
senten = input("Enter the text = ")
sentences = senten.split()
#sentences = [str(a) for a in sentences]
#sentences = eval(sentences)
#print(sentences)
#senten = raw_input("Hii ")
#sentences = list(map(str, senten.split()))
analyzer = SentimentIntensityAnalyzer()
for sentence in sentences:
vs = analyzer.polarity_scores(sentence)
print("{:-<65} {}".format(sentence, str(vs)))
import feedparser
from vaderSentiment import SentimentIntensityAnalyzer
import collections
d = feedparser.parse('http://rss.nzherald.co.nz/rss/xml/nzhrsscid_000000001.xml')
analyzer = SentimentIntensityAnalyzer()
if __name__ == '__main__':
result = {}
for post in d.entries:
print ("\nAnalyze:", post.title)
print ("Sentence:", post.description)
vs = analyzer.polarity_scores(post.description)
print("{:-<65} {}".format(post.title, str(vs)))
print("Compound:", vs['compound'])
result[vs['compound']] = post.title
keylist = result.keys()
keylist = sorted(result.keys())
print ("\n %s: %s" % (keylist[0], result[keylist[0]]))
def getNeautral(text):
analyzer = SentimentIntensityAnalyzer()
vs = analyzer.polarity_scores(text)
rating = vs["neu"]
return {'score': rating}
def getPositive(text):
analyzer = SentimentIntensityAnalyzer()
vs = analyzer.polarity_scores(text)
rating = vs["pos"]
return {'score': rating}
def getCompound(text):
analyzer = SentimentIntensityAnalyzer()
vs = analyzer.polarity_scores(text)
rating = vs["compound"]
return {'score': rating}
def getAll(text):
analyzer = SentimentIntensityAnalyzer()
vs = analyzer.polarity_scores(text)
return vs
# def fileRead(folder):
# for filename in os.listdir(folder):
# f = open(folder + "/" + filename, encoding="utf-8-sig")
# for l in f:
# l.strip()
# s = l.split("ред")
# file.extend(s)
# # fileRead("accident")
# fileRead("Test")
# print(len(file))
# b = input("Enter Bangla sentence :")
analyzer = SentimentIntensityAnalyzer()
t = analyzer.polarity_scores(
"If the speed of foreign aid increases and the increase in the revenue collection will reduce the amount of debt."
)
print(str(t))
file = open('data.txt', 'r', encoding="utf-8-sig")
file = [i.strip() for i in file]
for line in file:
line.strip()
# print(line)
# textBlobTrans = TextBlob(line)
# print(textBlobTrans)
# c = str(textBlobTrans.translate(to='en'))
# print(c)
# analyze = TextBlob(c)
# print('By TextBlob Sentiment analysis')
# print(analyze.sentiment)