Update column value based on value in other column

Question

I am referencing to this thread. I want to update a column value based on value in another column.

name_list = ['C', 'E']
df = pd.DataFrame([('A', 'buy'),
                  ('B', 'sell'),
                  ('C', 'hold'),
                  ('D', 'loan'),
                  ('E', 'hold')], columns=['name', 'action'])

    name     action
1    A          buy
2    B          sell           
3    C          hold
4    D          loan
5    E          hold

Here is two tries I came up with:

df['action'] = df.apply(lambda x: 'removed' if x['name'] in name_list else df['action'])
df['action'] = df.apply(lambda x: 'removed' if x['name'].isin(name_list) else df['action'])

Both tries above gives this error KeyError: 'name' The expected output is

    name     action
1    A          buy
2    B          sell           
3    C          removed
4    D          loan
5    E          removed

What am I doing wrong?

Answer 1

Simply use loc:

df.loc[df['name'].isin(name_list),'action']='removed'

OR

For you current method pass axis=1 but using apply() is slow for this simple case:

df['action'] = df.apply(lambda x: 'removed' if x['name'] in name_list else x['action'],axis=1)

output of df:

    name    action
0   A       buy
1   B       sell
2   C       removed
3   D       loan
4   E       removed

Note: you can also use np.where(): np.where(df['name'].isin(name_list),'removed',df['action'])

Answer 2

Use df.where-

df['action'] = df['action'].where(~df['name'].isin(name_list), 'removed')

Output

0        buy
1       sell
2    removed
3       loan
4    removed
Name: action, dtype: object