|
|
@@ -0,0 +1,67 @@ |
|
|
import re |
|
|
import json |
|
|
|
|
|
# save the positive words into a list called p_list |
|
|
with open('positive.txt') as f: |
|
|
p_txt = f.read() |
|
|
p_txt = re.sub('[,\.()":;!@#$%^&*\d]|\'s|\'', '', p_txt) |
|
|
p_list = p_txt.replace('\n',' ').replace(' ',' ').lower().split(' ') |
|
|
# test if cool is in the list |
|
|
print 'cool is in the postive list: ', 'cool' in p_list |
|
|
|
|
|
# save the negative words into a list called n_list |
|
|
with open('negative.txt') as f: |
|
|
n_txt = f.read() |
|
|
n_txt = re.sub('[,\.()":;!@#$%^&*\d]|\'s|\'', '', n_txt) |
|
|
n_list = n_txt.replace('\n',' ').replace(' ',' ').lower().split(' ') |
|
|
# test if abrade is in the list |
|
|
print 'abrade is in the negative list: ', 'abrade' in n_list |
|
|
# test if cool is in the list |
|
|
print 'cool is in the negative list: ', 'cool' in p_list |
|
|
|
|
|
# process the tweets |
|
|
with open('data.txt') as f: |
|
|
|
|
|
txt = f.read() |
|
|
txt = re.sub('[,\.()":;!@#$%^&*\d]|\'s|\'', '', txt) |
|
|
word_list = txt.replace('\n',' ').replace(' ',' ').lower().split(' ') |
|
|
|
|
|
# create empty dictionaries |
|
|
word_count_dict = {} |
|
|
word_count_positive = {} |
|
|
word_count_negative= {} |
|
|
|
|
|
for word in word_list: |
|
|
# count all words frequency |
|
|
if word in word_count_dict.keys(): |
|
|
word_count_dict[word] += 1 |
|
|
else: |
|
|
word_count_dict[word] = 1 |
|
|
# count if it is a positive word |
|
|
if word in p_list: |
|
|
if word in word_count_positive.keys(): |
|
|
word_count_positive[word] += 1 |
|
|
else: |
|
|
word_count_positive[word] = 1 |
|
|
# else see if it is a negative word |
|
|
elif word in n_list: |
|
|
if word in word_count_negative.keys(): |
|
|
word_count_negative[word] += 1 |
|
|
else: |
|
|
word_count_negative[word] = 1 |
|
|
else: # do nothing |
|
|
pass |
|
|
|
|
|
list_dict = sorted(word_count_dict.items(), key=lambda x:x[1], reverse=True) |
|
|
list_positive = sorted(word_count_positive.items(), key=lambda x:x[1], reverse=True) |
|
|
list_negative = sorted(word_count_negative.items(), key=lambda x:x[1], reverse=True) |
|
|
|
|
|
with open('word_count.csv', 'w')as f1: |
|
|
for i in list_dict: |
|
|
f1.write('%s,%s\n' %(i[0],str(i[1]))) |
|
|
with open('word_positive.csv', 'w')as f1: |
|
|
for i in list_positive: |
|
|
f1.write('%s,%s\n' %(i[0],str(i[1]))) |
|
|
with open('word_negative.csv', 'w')as f1: |
|
|
for i in list_negative: |
|
|
f1.write('%s,%s\n' %(i[0],str(i[1]))) |
octoparse created this gist