Data cleaning: regex to replace numbers

Question

I have this dataframe:

p=pd.DataFrame({'text':[2,'string']})

and trying to replace digit 2 by an 'a' using this code:

p['text']=p['text'].str.replace('\d+', 'a')

But instead of letter a and get NaN?

What am I doing wrong here?

Answer 1

In your dataframe, the first value of the text column is actually a number, not a string, thus the NaN error when you try to call .str. Just convert it to a string first:

p['text'] = p['text'].astype(str).str.replace('\d+', 'a')

Output:

>>> p
     text
0       a
1  string

(Note that .str.replace is soon going to change the default value of regex from True to False, so you won't be able to use regular expressions without passing regex=True, e.g. .str.replace('\d+', 'a', regex=True))