How to Groupby and calculate max based on filter in pandas dataframe-CodePudding

So, i have a dataframe like this:

I want to groupby based on Field1, if count is greater than 2, then find max Field2 and create a new field and set it to True.

I tried

import pandas as pd 
pd.read_csv("c:/test.csv")
df["Field3"] = df.groupby(["Field1"])["Field2"].transform("max")

But it didn't work.

CodePudding user response：

We have to do an additional transform to check the counts per Field1 which are greater than 2

g = df.groupby("Field1")["Field2"]
df['Field3'] = g.transform('count').gt(2) & df['Field2'].eq(g.transform('max'))

Alternatively you can also use the single transform with lambda function to check for the conditions but this might be slower that the first approach on larger dataframes

df['Field3'] = df.groupby("Field1")["Field2"].transform(
                         lambda s: (s == max(s)) * (len(s) > 2))

  Field1  Field2  Field3
0      a       3   False
1      a       5    True
2      a       3   False
3      b       2   False
4      c       1   False
5      b       6   False