Merge two columns by mapping column names to row values of another column-CodePudding

Is it possible to merge two columns by mapping column names to values in antoher column?

df = pd.DataFrame(
    {
        "id": [1, 1, 2, 3],
        "label": ["foo", "moo", "foo", "moo"],
        "foo.foo": ["foo", "foo", "foo", np.nan],
        "moo.moo": ["moo", "moo", np.nan, np.nan],
    }
)

expected result:

id label foo.moo 
1  foo   foo     
1  moo   moo     
2  foo   foo

CodePudding user response：

check if this works for you

df['foo_moo_updated'] = df['foo.moo'].mask(df.label=='foo',df['foo.foo'])
df[~df['foo_moo_updated'].isnull()]

    id  label   foo.foo foo.moo foo_moo_updated
0   1   foo foo moo foo
1   1   moo foo moo moo
2   2   foo foo NaN foo