def sentiment_words(self, filename):
"""
Parameters
----------
filename : str
file path for sentiment scores. Represented with a pos, neg, or both score. .
Returns
-------
Dictionary of sentiment scores for words.
"""
df = pd.read_table(filename, skiprows=26)
df['score'] = df['PosScore'] - df['NegScore']
df = df[['SynsetTerms', 'score']]
df.columns = ['words', 'score']
# remove neutral words
mask = df['score'] != 0
df = df[mask]
# Regex to find number
rx1 = re.compile('#([0-9])')
# Regex to find words
verEx = VerEx()
exp = verEx.range('a', 'z', 'A', 'Z')
rx2 = re.compile(exp.source())
sent_dict = {}
for i, row in df.iterrows():
w = row['words']
s = row['score']
nums = re.findall(rx1, w)
w = w.split(' ')
words = []
if len(w) == 1:
words = ''.join(re.findall(rx2, str(w)))
else:
words = [''.join(re.findall(rx2, str(string))) for string in w]
for nn, ww in zip(nums, words):
# only sentiment for the most common meaning of the word
if nn == '1':
sent_dict[ww] = s
return sent_dict
# In[198]:
from verbalexpressions import VerEx
# In[199]:
verEx = VerEx()
# In[200]:
strings = ['123Abdul233', '233Raja434', '223Ethan Hunt444']
# In[201]:
expression = verEx.range('a', 'z', 'A', 'Z', ' ')
# In[202]:
expression.source()
# In[204]:
import re
re_exp = expression.compile()
# In[205]:
re.findall(re_exp, strings[0])
