re.search if else statement and replace()

user_14426115 0
Viewed 175

I'm quite new to programming. I'm trying to create an if elif statement with re.search and the replace function, but don't get the result I want for the else statement. So this is what I want to do:

if re.search('.*[Cc]orona.*|.*[Qq]uarantine.*', str(table['reason'])):
    table['new_reason'] = table['reason'].replace('.*[Cc]orona.*|.*[Qq]uarantine.*', 'covid related', regex = True)
elif re.search('.*', (str(table['reason'])):
    table['new_reason'].replace('.*', 'other reason', regex = True)'

In other words, I'm trying to create a new column where the value would be "covid related" if the 'reason' column contains the word corona or quarantine. If not, I want it to return "other reason". Now, it only works with the if statement (i.e. transforms the values with 'corona' and 'quarantine' in them to 'covid related'). I does not transform all the other values to 'other reason'.

I also tried this, but get the same result:

if re.search('.*[Cc]orona.*|.*[Qq]uarantine.*', str(table['reason'])):
    table['new_reason'] = table['reason'].replace('.*[Cc]orona.*|.*[Qq]uarantine.*', 'covid related', regex = True)
else:
    table['new_reason'].replace('.*', 'other reason', regex = True)

Appreciate any help.

pandas if-statement python replace regex
1 Answers

Use .str.contains('quarantine|corona', case=False) (or, to match whole words, r'\b(?:quarantine|corona)\b') condition with np.where:

import pandas as pd
import numpy as np
df = pd.DataFrame({'reason':['__ Corona ___', '++++ quarantine +++', '__ CORONA ___', '++++ QUARANTINE +++', '--- NA ---']})
df['new_reason'] = np.where(df['reason'].str.contains('quarantine|corona', case=False), 'covid related', 'other reason')
>>> df
                reason     new_reason
0        __ Corona ___  covid related
1  ++++ quarantine +++  covid related
2        __ CORONA ___  covid related
3  ++++ QUARANTINE +++  covid related
4           --- NA ---   other reason
user_113298900
Related