Replace multiple characters from one column with NaN in Python

Question

I want to replace the position words from strings column: if they are either present sole or in multiple but join with , and space.

    id                         strings
0    1                           south
1    2                           north
2    3                            east
3    4                            west
4    5               west, east, south
5    6                      west, west
6    7                    north, north
7    8                    north, south
8    9  West Corporation global office
9   10                     West-Riding
10  11      University of West Florida
11  12                       Southwest

My expected result will like this. Please note if they are components of phrase or words, then I don't need to replace them.

Is it possible to do that? Thank you.

    id                         strings
0    1                             NaN
1    2                             NaN
2    3                             NaN
3    4                             NaN
4    5                             NaN
5    6                             NaN
6    7                             NaN
7    8                             NaN
8    9  West Corporation global office
9   10                     West-Riding
10  11      University of West Florida
11  12                       Southwest

The following code works, but I just wonder if there are some more concise methods?

df['strings'].astype(str).replace('south', np.nan).replace('north', np.nan)\
.replace('west', np.nan).replace('east', np.nan).replace('west, east', np.nan)\
.replace('west, west', np.nan).replace('north, north', np.nan).replace('west, east', np.nan)\
.replace('north, south', np.nan)

Answer 1

Using Regex.

Ex:

df = pd.DataFrame({'strings': ['south', 'north', 'east', 'west', 'west, east, south', 'west, west', 'north, north', 'north, south', 'West Corporation global office', 'West-Riding', 'University of West Florida', 'Southwest']})
df['R'] = df['strings'].replace(r"\b(south|north|east|west)\b,?", np.NAN, regex=True)
print(df)

Output:

                           strings                               R
0                            south                             NaN
1                            north                             NaN
2                             east                             NaN
3                             west                             NaN
4                west, east, south                             NaN
5                       west, west                             NaN
6                     north, north                             NaN
7                     north, south                             NaN
8   West Corporation global office  West Corporation global office
9                      West-Riding                     West-Riding
10      University of West Florida      University of West Florida
11                       Southwest                       Southwest

Answer 2

First use Series.str.split, forward filling for replace missing values, test if all matched values by DataFrame.isin and DataFrame.all for mask and last set missing values by Series.mask:

L = ['south','north','east','west']
m = df['strings'].str.split(', ', expand=True).ffill(axis=1).isin(L).all(axis=1)

df['strings'] = df['strings'].mask(m)
print (df)
    id                         strings
0    1                             NaN
1    2                             NaN
2    3                             NaN
3    4                             NaN
4    5                             NaN
5    6                             NaN
6    7                             NaN
7    8                             NaN
8    9  West Corporation global office
9   10                     West-Riding
10  11      University of West Florida
11  12                       Southwest

Another idea with sets, isdisjoint and Series.where:

m = [set(x.split(', ')).isdisjoint(L) for x in df['strings']]
df['strings'] = df['strings'].where(m)
print (df)
    id                         strings
0    1                             NaN
1    2                             NaN
2    3                             NaN
3    4                             NaN
4    5                             NaN
5    6                             NaN
6    7                             NaN
7    8                             NaN
8    9  West Corporation global office
9   10                     West-Riding
10  11      University of West Florida
11  12                       Southwest