TypeError: expected string or bytes-like object on Pandas using Fuzzy matching

Question

Background

I have a df

import pandas as pd
import nltk
from fuzzywuzzy import fuzz
from fuzzywuzzy import process

df= pd.DataFrame({'ID': [1,2,3], 
                           'Text':['This num dogs and cats is (111)888-8780 and other',
                              'dont block cow 23 here',
                              'cat two num: dog  and cows here']    
                      })

I also have a list

 word_list = ['dog', 'cat', 'cow']

and a function that is supposed to do fuzzy matching on the Text column of the df with the word_list

def fuzzy(row, word_list):
    
    tweet = row[0]
    fuzzy_match = []

    for word in word_list:
     
        token_words = nltk.word_tokenize(tweet)
        
        for token in range(0, len(token_words) - 1):
                
            fuzzy_fx = process.extract(word_list[word], token_words[token], limit=100, scorer = fuzz.ratio)
            fuzzy_match.append(fuzzy_fx[0])

    return pd.Series([fuzzy_match], index = ['Fuzzy_Match'])

I then join

df_fuzz = df.join(df.apply(lambda x: fuzzy(x, word_list), axis = 1))

But I get an error

TypeError: expected string or bytes-like object

Desired output My desired output would be a 1) new column Fuzzy_Match with the output of the fuzzy function

    ID  Text                                                 Fuzzy_Match
0   1   This num dogs and cats is (111)888-8780 and other   output of fuzzy 1
1   2   dont block cow 23 here                              output of fuzzy 2
2   3   cat two num: dog and cows here                      output of fuzzy 3

Question What do I need to do to get my desired output?

Answer 1

This should work:

In [32]: def fuzzy(row, word_list):
    ...:     tweet = row[1]
    ...:     fuzzy_match = []
    ...:     token_words = nltk.word_tokenize(tweet) 
    ...:     for word in word_list:
    ...: 
    ...:         fuzzy_fx = process.extract(word, token_words, limit=100, scorer = fuzz.ratio)
    ...:         fuzzy_match.append(fuzzy_fx[0])
    ...: 
    ...:     return pd.Series([fuzzy_match], index = ['Fuzzy_Match'])

df_fuzz = df.join(df.apply(lambda x: fuzzy(x, word_list), axis = 1))

process.extract() expects a list as the second argument. you can read more about it here. python fuzzywuzzy's process.extract(): how does it work?