def main():
choice='TP'
prepare_dictionary();
if(choice == 'TP'):
global pathVars;
global pathToAuthors;
global gloveText;
for i, item in enumerate(pathVars):
pathToFold = os.getcwd() + "/data/folds"+ "/" + pathTopics[i]+ "_folds";
reorderedData = reorder(readData.readData(pathVars[i], pathToAuthors, pathTopics[i]), pathToFold);
learn(reorderedData, pathTopics[i]);
elif(choice == 'E'):
sentence = raw_input("Enter your sentence..");
topic = raw_input(' What is the topic? ');
predict(sentence.lower(), topic);
def main():
choice = 'TP'
prepare_dictionary()
if (choice == 'TP'):
global pathVars
global pathToAuthors
global gloveText
for i, item in enumerate(pathVars):
pathToFold = os.getcwd(
) + "/data/folds" + "/" + pathTopics[i] + "_folds"
reorderedData = reorder(
readData.readData(pathVars[i], pathToAuthors, pathTopics[i]),
pathToFold)
learn(reorderedData, pathTopics[i])
elif (choice == 'E'):
sentence = raw_input("Enter your sentence..")
topic = raw_input(' What is the topic? ')
predict(sentence.lower(), topic)
def update_output_div(n, tag):
rows = [[learn.predict(MODEL, x, TOKENS),
x.replace('\n', '')] for x in twitter.get_last_tweets(tag)]
emos = [row[0] for row in rows]
x = ['Neutral', 'Negative', 'Positive']
y = [emos.count('neutral'), emos.count('negative'), emos.count('positive')]
fig = go.Figure(data=[go.Pie(labels=x, values=y)],
layout=go.Layout(title='Points Accumulation',
showlegend=False))
return fig
def testAlgo(name, classifier, iterations, output=None):
print name
expected = map(learn.getClassification, tests)
testCount = len(expected)
scores = []
for iteration in range(0, iterations):
actual = learn.predict(classifier, trainings, tests)
scores.append(getScore(expected, actual))
# print scores
average = numpy.mean(scores)
percentage = int((float(average) / float(testCount)) * 100.0)
# print str(average) + "/" + str(float(totalCount)) + "\n"
print str(percentage) + '%'
def update_output_div(n, tag):
rows = [[learn.predict(MODEL, x, TOKENS),
x.replace('\n', '')] for x in twitter.get_last_tweets(tag)]
return [generate_table(rows)]
trainings = json.load(data_file)
with open(test_filename) as data_file:
tests = json.load(data_file)
def getScore(expected, actual):
numCorrect = 0
length = len(expected)
for index in range(0, length):
if (expected[index] == actual[index]):
numCorrect += 1
return numCorrect
expected = map(learn.getClassification, tests)
actual = learn.predict(classifier, trainings, tests)
def testAlgo(name, classifier, iterations, output=None):
print name
expected = map(learn.getClassification, tests)
testCount = len(expected)
scores = []
for iteration in range(0, iterations):
actual = learn.predict(classifier, trainings, tests)
scores.append(getScore(expected, actual))
# print scores
average = numpy.mean(scores)
percentage = int((float(average) / float(testCount)) * 100.0)
# print str(average) + "/" + str(float(totalCount)) + "\n"
print str(percentage) + '%'