Introduction

I’ll show how to assign new value for each unique group in Pandas DataFrame.

import pandas as pd

data = {
    "ID": [1, 1, 1, 2, 2, 3, 3],
    "class": ["Lower", "Moderate", "Moderate", "Lower", "High", "High", "Lower"]
}

df = pd.DataFrame(data)

df

	ID	class
0	1	Lower
1	1	Moderate
2	1	Moderate
3	2	Lower
4	2	High
5	3	High
6	3	Lower

Now, let assign new values according to class column and then rearrange class column in order to assign new group names.

value_map = {"Lower": 33, "Moderate": 34, "High": 40}

df["new"] = df["class"].map(value_map)

df

	ID	class	new
0	1	Lower	33
1	1	Moderate	34
2	1	Moderate	34
3	2	Lower	33
4	2	High	40
5	3	High	40
6	3	Lower	33

df["Max"] = (df.set_index("class")
                .groupby("ID")["new"]
                .transform("idxmax").values)

df

	ID	class	new	Max
0	1	Lower	33	Moderate
1	1	Moderate	34	Moderate
2	1	Moderate	34	Moderate
3	2	Lower	33	High
4	2	High	40	High
5	3	High	40	High
6	3	Lower	33	High

df["Max"] = df["Max"].mask(df["ID"].duplicated(), "")

df

	ID	class	new	Max
0	1	Lower	33	Moderate
1	1	Moderate	34
2	1	Moderate	34
3	2	Lower	33	High
4	2	High	40
5	3	High	40	High
6	3	Lower	33