Return total count of list of words in a string python?

user_12458212 0
Viewed 91

I know the following function will simply return a count of a word in a sentence. I'm trying to loop through a df['col'] using a list of words. I currently have my function set up to loop over each row. But I'd like to know if there's a way to avoid using a nested loop.

def count_occurrences(word, sentence):
    return sentence.lower().split().count(word)

For example, say my column is as below. I want to count how many times the words in the following list occurs in each sentence e.g.

list_of_words = ['green','apple']

ind |   df['col']
------------------------------------------
1.  |  'green apple red apple blue apple'
2.  |  'green apple green apple green apple'

Output:

[4,6]
pandas python string
3 Answers

You can try with pandas.Series.str.count:

>>> df
                                   col
0     green apple red apple blue apple
1  green apple green apple green apple
>>>
>>> df.col.str.count('|'.join(list_of_words)).tolist()
[4, 6]
user_144210350

try this:

df_counts = df["col"].str.split().explode().value_counts()
df_counts[list_of_words]

output:

green    4
apple    6
Name: col, dtype: int64

This may be a more useful output form than a list of counts. Or, if you just want the list:

df_counts[list_of_words].to_list()
user_50329410

It will be better adding /b in case you have pineapple 

df['cnt'] = df['col'].str.count(r'/b|/b'.join(list_of_words))
user_79645270
Related