[FIXED] How to remove duplicates in each row

Issue

I have a dataframe with duplicates in each row. I would like to return the non duplicates with a new column name like Num1 thru Num5. For instance,

import numpy as np
import pandas as pd
 
df1 = pd.DataFrame([[1,1,2,4,5,6,7,7],
           [2,5,6,7,22,23,34,7],
           [3,3,5,6,7,45,46,7],
           [4,6,7,14,29,32,33,7],
           [5,6,7,13,23,33,35,7],
           [6,1,6,7,8,9,10,7],
           [7,0,2,5,7,10,30,7]], 
           columns = ['Row_Num', 'Num1','Num2','Num3','Num4','Num5','Num6','Num7'])

I would like my results to be like this.

result = pd.DataFrame([[1,1,2,4,5,6],
           [2,5,6,22,23,34],
           [3,3,5,6,45,46],
           [4,6,14,29,32,33],
           [5,6,13,23,33,35],
           [6,1,6,8,9,10],
           [7,0,2,5,10,30]], 
           columns = ['Row_Num', 'Num1','Num2','Num3','Num4','Num5'])

Solution

To avoid transposing the dataframe you can use iterrows instead of items as in my other answer, and using drop_duplicates as @Timeless proposed this is a bit faster than the other answers currently:

remaining_values = {row_num: values.drop_duplicates(keep=False).values
                    for row_num, values in df1.set_index('Row_Num').iterrows()}
result = pd.DataFrame(remaining_values).T.rename(columns=lambda i: f"Num{i+1}")
print(result)

Output:

   Num1  Num2  Num3  Num4  Num5
1     1     2     4     5     6
2     5     6    22    23    34
3     3     5     6    45    46
4     6    14    29    32    33
5     6    13    23    33    35
6     1     6     8     9    10
7     0     2     5    10    30

The last step will only work if there are an equal number of duplicated values in each row, as others have pointed out.

Answered By - Bill