I think you need:
df['features'] = df.values.tolist()
print(df)
Atr1 Atr2 features
0 1 A [1, A]
1 2 B [2, B]
2 4 C [4, C]
If you have multiple columns and want to select particular columns then:
df = pd.DataFrame({"Atr1":[1,2,4],"Atr2":['A','B','C'],"Atr3":['x','y','z']})
print(df)
Atr1 Atr2 Atr3
0 1 A x
1 2 B y
2 4 C z
#Selecting Atr2 and Atr3 columns
df['features'] = df[['Atr2','Atr3']].values.tolist()
print(df)
Atr1 Atr2 Atr3 features
0 1 A x [A, x]
1 2 B y [B, y]
2 4 C z [C, z]
1
solved How can I create a vector in pandas dataframe from other attributes of the dataframe?