CODEWITHSUNDEEP
python-3.xpandas
Kalyan
Kalyan

Reputation: 1940

Getting Series' object has no attribute 'split'", 'occurred at index id when removing frequent word from tweets

I am trying sentiment analysis from the twitter post. I am new to sentiment analysis.In text preprocessing phase, I have encountered a problem to remove frequent words from twits. i want to remove most frequent words from twits so I have counted most frequent terms in twit by

freq=pd.Series(''.join(traindata['tweet']).split()).value_counts()[:10]

then i have converted the freq series into list

freq=list(freq.index)

Up to this point, my result is showingenter image description here

For filtering my twitter_word column by removing frequently used words. I have used below code

traindata['tweet']=traindata.apply(lambda x:" ".join(x for x in x.split() if x not in freq))

and I have got below error

File "C:\Users\codemen\Anaconda3\lib\site-packages\pandas\core\generic.py", line 3614, in __getattr__ return object.__getattribute__(self, name) AttributeError: ("'Series' object has no attribute 'split'", 'occurred at index id')

kindly help me to figure out the problem. Thank you

Upvotes: 1

Views: 8998

Answers (1)

jezrael
jezrael

Reputation: 863351

I believe you need specify column for apply, else looping all columns of DataFrame:

f = lambda x:" ".join(x for x in x.split() if x not in freq)
traindata['tweet'] = traindata['tweet'].apply(f)

Upvotes: 0

Related Questions