Formatting the scientific notation phone number in python

Question

I have my phone_number column listed below.

phone_number
--------------
001 1234567890
380 1234567890
 27 1234567890
001 +11234567890
2.56898E+11
1 1234567890
123-456-7890
 +1 (123) 456-7890
(123) 456-7890
NaN

The following step worked fine

character = '[^0-9]+'
df.phone_number.str.replace(character, '')

The result I got is

phone_number
--------------
11234567890
3.80123E+12
2.71234E+11
11234567890
2.56898E+11
11234567890
1234567890
11234567890
1234567890
NaN

Is there any elegant way to deal with the scientific notation format? I want them to be 11234567890 or longer because of the country code. From there I think I can figure out how to get both international and the US phone number formats. Thanks in advance.

Answer 1

You can use conversion to Int64 / string dtypes:

s1 = (pd.to_numeric(df['phone_number'], errors='coerce')
        .astype('Int64').astype('string')
      )

s2 = df['phone_number'].str.replace(r'\D+', '', regex=True)

df['phone_number_clean'] = s1.fillna(s2)

print(df)

Output:

        phone_number phone_number_clean
0     001 1234567890      0011234567890
1     380 1234567890      3801234567890
2      27 1234567890       271234567890
3   001 +11234567890     00111234567890
4        2.56898E+11       256898000000
5       1 1234567890        11234567890
6       123-456-7890         1234567890
7  +1 (123) 456-7890        11234567890
8     (123) 456-7890         1234567890

Note that depending on the float precision and the way the number was converted to scientific notation, you might lose important digits.