Replacing backtick in pandas-CodePudding

I have the following data frame:

my_places = ['Bernardo O\'Higgins', 'Bernardo O`Higgins', 'Bernardo O\"Higgins'] 
adf = pd.DataFrame({'my_places':my_places})
adf
    my_places
0   Bernardo O'Higgins
1   Bernardo O`Higgins
2   Bernardo O"Higgins

And I try to replace the characters ',` and " with ''. So I tried this:

adf['my_places_check'] = adf['my_places'].replace(''|"|`','')

But I get an error:

SyntaxError: EOL while scanning string literal

I think I have to 'escape' the special characters, so I tried this:

adf['my_places_check'] = adf['my_places'].replace('\'|\"|\`','')

But it returns this:

    my_places   my_places_check
0   Bernardo O'Higgins  Bernardo O'Higgins
1   Bernardo O`Higgins  Bernardo O`Higgins
2   Bernardo O"Higgins  Bernardo O"Higgins

And I expected:

    my_places   my_places_check
0   Bernardo O'Higgins  Bernardo OHiggins
1   Bernardo O`Higgins  Bernardo OHiggins
2   Bernardo O"Higgins  Bernardo OHiggins

Any help will be greatly appreciated.

CodePudding user response：

You need to use

adf['my_places_check'] = adf['my_places'].replace("""["`']""", '', regex=True)

Details:

"""["`']""" - a regex that is ["`']
'' - is the replacement (an empty string)
regex=True - make sure the pattern is parsed as a regex

If you use " or ' as a string literal delimiter, you will need to escape the corresponding literal char inside the string literal, so using a triple-quoted string literal is a more readable solution.