如何根据多个条件随机select行

Question

I have two datasets with 20 rows each. I am looking to randomly select 10 rows from each dataset following the criteria below.

df1 group:

• 8 mammals
• 2 reptiles

df2 group:

• 4 mammals
• 2 birds
• 3 reptiles
• 1 fish

6 terrestrial and 4 aquatic ecosystems for both

df1.query("Class = Mammal").sample(n=8)

df1.query("Class = Reptile").sample(n=2)

I've seen solutions like this that should work, but I can't figure out how to include the ecosystems requirement. AKA I want 8 mammals and 2 reptiles selected from group 1, ensuring that 6 of them come from terrestrial ecosystems and 4 from aquatic. I think there should be a way to do this with a groupby function of the two columns, but I haven't yet figured that out.

Sample data:

Common name	Class	Ecosystem
Lion	Mammal	Terrestrial
Humpback whale	Mammal	Aquatic
Crocodile	Reptile	Aquatic

Answer 1

I don't know how to do it in a clean way with just the built-in pandas functions like groupby . That said, here's a solution using random and lists.

import random

animal_class = ["Mammal"] * 8 + ["Reptile"] * 2

ecosystem = ["Terrestrial"] * 6 + ["Aquatic"] * 4
random.shuffle(ecosystem)  # randomly shuffle ecosystem

df1_selected = pd.DataFrame(columns=df1.columns)
for i in range(10):
    df1_selected = df1_selected.append(
        df1.query(f"Class = {animal_class[i]} and Ecosystem = {ecosystem[i]}").sample(n=1)
    )

Just change the animal_class to do the same thing for df2.