Input dataframe is as follows:
import pandas as pd
df_input = pd.DataFrame([[1,0,0,0,0], [0,1,0,0,0], [0,0,1,0,0],[0,0,0,1,0],[0,0,0,0,1]], columns=["A", "B","C","D","E"])
For all the rows, only one column has non zero entry. Accordingly, the output dataframe I am expecting is as follows:-
df_output=pd.DataFrame(['A','B','C','D','E'],columns=['Alphabet'])
The alphabet column should have the column name of non zero value. Please suggest
You can use a dot product after getting a boolean df checking for not equals 0.
If there are chances of multiple columns having a non zero value for a row and you want to get the first:
df_input.ne(0).dot(df_input.columns+',').str.split(",").str[0].to_frame('Alphabet')
If there will be only 1 column with a non zero value always , then we can use rstrip like @Shubham mentioned in their coments.
df_input.ne(0).dot(df_input.columns+',').str.rstrip(',').to_frame('Alphabet')
Alphabet
0 A
1 B
2 C
3 D
4 E
You can use idxmax along axis=1:
df_input.ne(0).idxmax(1).to_frame('Alphabet')
Alphabet
0 A
1 B
2 C
3 D
4 E
