def span(common_words):
top_tweets = []
final_list = []
getSentiment = act.load_classifier()
with open('uploads/TWEETS.csv', 'r', encoding="utf8") as wash:
tReader = csv.reader(wash, delimiter=',', quotechar='|')
csv.field_size_limit(20000000)
for i in tReader:
top_tweets.append(i)
for g in top_tweets:
if len(g) > 5 and common_words in g[5]:
text = act.processTweet2(g[5])
final = act.postprocess(text)
sentTweets = act.get_sentiment(final, getSentiment)
final_list.append([g[1], final, sentTweets])
if (len(final_list) == 10):
break
date = []
for span in top_tweets:
if len(span) > 5 and common_words in span[5]:
date.append(span[1])
if (len(date) == len(top_tweets)):
break
Fdate = date[1:2]
Bdate = date[-1:]
return [Fdate, Bdate]
def class_dataB_test(clf="Support Vector Machine"):
ts = []
classifier = fact.load_classifier(clf)
with open('uploads/dataB.csv', 'r', encoding="utf8") as wash:
tReader = csv.reader(wash, delimiter=',', quotechar='|')
csv.field_size_limit(20000000)
for i in tReader:
ts.append(i)
gPlot={'jan':[0,0,0],'feb':[0,0,0],'mar':[0,0,0],'april':[0,0,0],'oct':[0,0,0],'jun':[0,0,0],'jul':[0,0,0],'may':[0,0,0],'aug':[0,0,0],'sep':[0,0,0],'nov':[0,0,0],'dec':[0,0,0]}
monthtocheck=['jan','feb','mar','april','oct','jun','jul','may','aug','sep','nov','dec']
postweets = 0
negtweets = 0
for t in ts:
# text = fact.processTweet(t)
if len(t) > 5:
text = t[5]
try:
#month = text.split(',')[1].split(' ')[1]
month = t[1]
month = month.lower()
print(month)
sentiment = fact.get_sentiment(text, classifier)
if(sentiment == "positive"):
if month in monthtocheck:
gPlot[month][0] += 1
postweets+=1
elif(sentiment == "negative"):
if month in monthtocheck:
gPlot[month][1] += 1
negtweets+=1
except IndexError:
continue
positives = []
negatives = []
months = ['jan', 'feb', 'mar', 'april', 'may', 'jun', 'jul', 'aug', 'sep', 'oct', 'nov', 'dec']
for month in months:
positives.append(gPlot[month][0])
negatives.append(gPlot[month][1])
#print(positives, negatives)
return positives, negatives
# print('\n')
# class_dataB()
def test(testset=testset, weight_file=args.weights):
"""tests classifier on name->gender
Args:
train: % of examples to train with (e.g., 0.8)
"""
start = time.time()
classifier = load_classifier(weight_file)
print("Testing Naive Bayes Classifer on %d examples (%s)" %
(len(testset), time_since(start)))
testset = apply_features(_gender_features, testset, labeled=True)
acc = nltk.classify.accuracy(classifier, testset)
print("Testing accuracy is %.2f%% on %d examples (%s)" %
(acc * 100, len(testset), time_since(start)))
return acc
def class_dataB(clf="Support Vector Machine"):
ts = compB()
classifier = fact.load_classifier(clf)
sen = []
gPlot={'jan':[0,0,0],'feb':[0,0,0],'mar':[0,0,0],'april':[0,0,0],'oct':[0,0,0],'jun':[0,0,0],'jul':[0,0,0],'may':[0,0,0],'aug':[0,0,0],'sep':[0,0,0],'nov':[0,0,0],'dec':[0,0,0]}
monthtocheck=['jan','feb','mar','april','oct','jun','jul','may','aug','sep','nov','dec']
postweets = 0
negtweets = 0
for t in ts:
text = fact.processTweet(t)
try:
month = text.split(',')[1].split(' ')[1]
month = month.lower()
sentiment = fact.get_sentiment(text, classifier)
if(sentiment == "positive"):
if month in monthtocheck:
gPlot[month][0] += 1
postweets+=1
elif(sentiment == "negative"):
if month in monthtocheck:
gPlot[month][1] += 1
negtweets+=1
except IndexError:
continue
posData = []
negData = []
months = ['jan', 'feb', 'mar', 'april', 'may', 'jun', 'jul', 'aug', 'sep', 'oct', 'nov', 'dec']
for month in months:
posData.append(gPlot[month][0])
negData.append(gPlot[month][1])
print(posData, negData)
pe = (100 * postweets / len(ts))
ne = (100 * negtweets / len(ts))
p = format(pe, '.2f')
n = format(ne, '.2f')
return posData, negData, p, n
def caTe(categories="categories"):
confirm = []
classifier = act.load_classifier()
holdOn = []
with open('uploads/TWEETS.csv', 'r', encoding="utf8") as wash:
tReader = csv.reader(wash, delimiter=',', quotechar='|')
csv.field_size_limit(20000000)
for i in tReader:
confirm.append(i)
for conF in confirm:
if len(conF) > 5 and categories in conF[5]:
text = act.processTweet2(conF[5])
sentiment = act.get_sentiment(text, classifier)
holdOn.append([text, sentiment])
if len(holdOn) == 10:
break
return holdOn
def viewMore(monthData, sentData):
more = []
classifier = act.load_classifier()
collectData = []
with open('uploads/TWEETS.csv', 'r', encoding="utf8") as wash:
tReader = csv.reader(wash, delimiter=',', quotechar='|')
csv.field_size_limit(20000000)
for i in tReader:
more.append(i)
for vMore in more:
if len(vMore) > 5 and monthData in vMore[1]:
text = act.processTweet2(vMore[5])
sentiment2 = act.get_sentiment(text, classifier)
if (sentiment2 == sentData):
collectData.append([vMore[1], text, sentiment2])
if len(collectData) == 10:
break
return collectData
def classify__(clf="Support Vector Machine"):
ts = getTweetFromFile()
getIt = len(ts)
classifier = act.load_classifier(clf)
avgPos = []
avgNeg = []
gPlot = {
'jan': [0, 0, 0],
'feb': [0, 0, 0],
'mar': [0, 0, 0],
'april': [0, 0, 0],
'oct': [0, 0, 0],
'jun': [0, 0, 0],
'jul': [0, 0, 0],
'may': [0, 0, 0],
'aug': [0, 0, 0],
'sep': [0, 0, 0],
'nov': [0, 0, 0],
'dec': [0, 0, 0]
}
monthtocheck = [
'jan', 'feb', 'mar', 'april', 'oct', 'jun', 'jul', 'may', 'aug', 'sep',
'nov', 'dec'
]
postweets = 0
negtweets = 0
neutral = 0
for t in ts:
text = act.processTweet(t)
try:
month = text.split(',')[1].split(' ')[1]
month = month.lower()
sentiment = act.get_sentiment(text, classifier)
if sentiment == "positive":
if month in monthtocheck:
gPlot[month][0] += 1
postweets += 1
elif sentiment == "negative":
if month in monthtocheck:
gPlot[month][1] += 1
negtweets += 1
else:
if month in monthtocheck:
gPlot[month][2] += 1
neutral += 1
except IndexError:
continue
# for sent in sentiment:
# if sentiment == "neutral":
# sen.append(sent)
# print(gPlot)
# print("Positive tweets percentage: {} %".format(100 * postweets / len(ts)))
# print("Negative tweets percentage: {} %".format(100 * negtweets / len(ts)))
# print("Neutral tweets percentage: {} %".format(100 * (len(ts) - negtweets - postweets) / len(ts)))
p = (100 * postweets / len(ts))
n = (100 * negtweets / len(ts))
ne = (100 * (len(ts) - negtweets - postweets) / len(ts))
if (p < n):
print("Negative")
elif (p > n):
print("Positive")
else:
print("Neutral")
pe = format(p, '.2f')
nee = format(n, '.2f')
neu = format(ne, '.2f')
positives = []
negatives = []
neutrals = []
months = [
'jan', 'feb', 'mar', 'april', 'may', 'jun', 'jul', 'aug', 'sep', 'oct',
'nov', 'dec'
]
for month in months:
positives.append(gPlot[month][0])
negatives.append(gPlot[month][1])
neutrals.append(gPlot[month][2])
for ee, gg in gPlot.items():
avgPos.append(gg[0])
avgNeg.append(gg[1])
aPos = sum(avgPos) / len(ts)
aNeg = sum(avgNeg) / len(ts)
avPos = format(aPos, '.4f')
avNeg = format(aNeg, '.4f')
return pe, nee, neu, positives, negatives, neutrals, avPos, avNeg, getIt