Python Pandas: Going through a list of cycles and making point of interest

August 30, 2022

To explain my problem easier I have created a dataset:

data = {'Cycle': ['Set1', 'Set1', 'Set1', 'Set2', 'Set2', 'Set2', 'Set2'],
        'Value': [1, 2.2, .5, .2,1,2.5,1]}

I want to create a loop that goes through the "Cycle" column and marks the max of each cycle with the letter A and the min with letter B, to output something like this:

POI = {'Cycle': ['Set1', 'Set1', 'Set1', 'Set2', 'Set2', 'Set2', 'Set2'],
        'Value': [1, 2.2, .5, .2,1,2.5,1],
         'POI': [0, 'A','B','B',0,'A',0]}

df2 = pd.DataFrame(POI)

I am new to Python, so as much detail as possible would be very helpful, as well as I am not exactly sure how to go through each cycle on its own to get these values, so explaining that would be great.

Thanks

>Solution :

Using numpy.select and groupby.transform:

g = df.groupby('Cycle')['Value']
df['POI'] = np.select([df['Value'].eq(g.transform('min')),
                       df['Value'].eq(g.transform('max'))],
                      ['A', 'B'])

# if you want 0 as default value (not '0')
df['POI'] = df['POI'].replace('0', 0)

output:

  Cycle  Value POI
0  Set1    1.0   0
1  Set1    2.2   B
2  Set1    0.5   A
3  Set2    0.2   A
4  Set2    1.0   0
5  Set2    2.5   B
6  Set2    1.0   0