I'm trying to sample 1000 unique users within a data. These can be any 1000 users. But I want to extract all rows for the 1000 unique users.

Input

from random import sample
df['User_ID'].sample(n=1000, random_state=1)

I tried the above code, but this just gives the unique IDs and not all rows for 1000 unique users.

CodePudding user response：

IIUC, get the unique values, sample and slice with isin and boolean indexing:

from random import sample

out = df[df['User_ID'].isin(random.sample(list(df['User_ID'].unique()), 1000))]