def compute_sentiment_score_byAnew(wordsList):
arousalScore = 0.0
valenceScore = 0.0
for word in wordsList:
if anew.exist(word):
arousalScore += anew.arousal(word)
valenceScore += anew.valence(word)
return (arousalScore, valenceScore)
def compute_sentiment_score_byAnew(wordsList):
arousalScore = 0.0
valenceScore = 0.0
for word in wordsList:
if anew.exist(word):
arousalScore += anew.arousal(word)
valenceScore += anew.valence(word)
return (arousalScore,valenceScore)
def stringfunction(string):
result = []
valence = []
arousal = []
k = 1
line = ""
for i in string:
#break sentences into paragraphs
if i not in (".", "!", "?"):
line = "".join([line, i])
else:
k == k + 1
#feed each sentence through the ANEW scorer
#First check to see that each line has at least 2 terms
term_list = line.split()
total = 0
for j in term_list:
if anew.exist(j) == True:
total = total + 1
if total > 1:
#At least 2 registering terms, score the line!
#return arousal with the highest valence!
grFiveV = False
grFiveA = False
toBeat = 0
key = anew.sentiment(term_list)
#finds distance value is from 5, puts into lisis
#retains +/- from center (5)
if key['valence'] > 5:
grFiveV = True
else:
grFiveV = False
if key['arousal'] > 5:
grFiveA = True
else:
grFiveA = False
#how far from 5
valence.append(key['valence'])
arousal.append(key['arousal'])
line = ""
#takes max list value and puts into result list.
#choose biggest valence away from 5
goal = 0
current1 = 0
for l in valence:
if abs(l - 5) > goal:
current1 = l
#repeat for arousal
goal = 0
current2 = 0
for n in arousal:
if abs(n - 5) > goal:
current2 = n
result.append(current1) #valence
result.append(current2) #arousal
return result
#iterating over term vectors, obtaining sentiment valence and arousal if more than two terms exist in the dictionary
print("Cleaning and scoring")
#keeping track of the vid column
vid = -1
for item in text_noasc:
termlist = item.split()
clean_terms = []
vid += 1
#remove the stopwords
for words in termlist:
if words not in stopwords:
clean_terms.append(words)
if sum(anew.exist(clean_terms)) >= 2:
valence.append((anew.sentiment(clean_terms))['valence'])
arousal.append((anew.sentiment(clean_terms))['arousal'])
scored_terms.append([term for term in clean_terms if anew.exist(term)])
video_ids.append((data['vid'][vid]).split()[:2])
#cleaning up the candidate names
video_ids = [' '.join(vid) for vid in video_ids]
#cleaning up the terms
scored_terms = [','.join(s_term) for s_term in scored_terms]
#combining the list of terms with their scores
values = zip(scored_terms, valence, arousal, video_ids)
#writing file
d = punc.sub( '', d )
term_vec.append( word_tokenize( d ) )
print len(term_vec)
# term_list = term_vec[100]
# print anew.exist(term_list)
# print anew.sentiment(term_list)
valence=[]
arousal=[]
sentiment_list=[]
for i in term_vec:
if sum(anew.exist(i)) >=2:
sentiment_list.append(i)
valence.append ((anew.sentiment(i))['valence'])
arousal.append ((anew.sentiment(i))['arousal'])
values = zip(sentiment_list, valence, arousal, email_id)
print ('Writing file')
headers = ['sentiment list','valence', 'arousal', 'ID']
with open('email_scores.csv', 'wb') as out:
writer = csv.writer( out )
#write headers
writer.writerow(headers)
#write the csv file
for value in values:
writer.writerow(value)