[FIXED] Pivoting based on multiple columns and rearrange data in Python Dataframe

Issue

have a python dataframe.

df1

Load           Instance_name            counter_name         counter_value
0              A                         bytes_read            0
0              A                         bytes_written         90
0              B                         bytes_read            100
0              B                         bytes_Written          90

1              A                         bytes_read            10
1              A                         bytes_written         940
1              B                         bytes_read            1100
1              B                         bytes_written         910

To simplify view, I need something like below ex. transform counter_name column fields to columns and rearrange data.***

df2 = 
Load     Instance_name  bytes_read     bytes_written   
0              A            0                    90       
0              B           100                   90
1              A            10                   940    
1              B           1100                  910

I am new to python dataframe libraries and not sure the right way to achieve this.

Solution

Try this:

df.set_index(['Load', 'Instance_name', 'counter_name'])['counter_value'].unstack()

Output:

counter_name  Load Instance_name  bytes_Written  bytes_read  bytes_written
0                0             A            NaN         0.0           90.0
1                0             B           90.0       100.0            NaN
2                1             A            NaN        10.0          940.0
3                1             B            NaN      1100.0          910.0

Note typo in input dataframe bytes_Written and bytes_written.

Answered By - Scott Boston