Add a new column in dataframe with common value-CodePudding

I have dataframe df1 as:

col1	col2	col3
Apple	Apple	Apple
orange	0	orange
Cake	0	0
0	Banana	0
0	grape	grape

Logic: To add the common value to a new column. If value (other than zero) is present in one or two columns then compare them and add the common value.

I want to add a new column (New_col) to this dataframe with values like below:

col1	col2	col3	New_col
Apple	Apple	Apple	Apple
orange	0	orange	orange
Cake	0	0	Cake
0	Banana	0	Banana
0	grape	grape	grape

Any suggestion how to do this? Thank you in advance!

CodePudding user response：

Assuming, you want to get the first non zero value per row, you can mask the zeros and bfill then get the first column:

df['NewCol'] = df.mask(df.eq('0')).bfill(axis=1).iloc[:,0]

NB. I am also assuming 0 is a string. You can now adapt this code to your needs

Output:

     col1    col2    col3  NewCol
0   Apple   Apple   Apple   Apple
1  orange       0  orange  orange
2    Cake       0       0    Cake
3       0  Banana       0  Banana

CodePudding user response：

An alternative to @mozway 's solution: Get the index of the first occurrence different from '0' with .ne("0") and .idxmax for each row using .apply, and assign the result to the new column.

>>> df["new_col"] = df.apply(lambda x: x[x.ne("0").idxmax()], axis=1)
>>> df

     col1    col2    col3  newCol
0   Apple   Apple   Apple   Apple
1  Orange       0  Orange  Orange
2    Cake       0       0    Cake
3       0  Banana       0  Banana