My attempt to simulate the retirement of images from a Zooniverse project given a retirement limit (variable = limit) and the number of subjects (variable = subjects).

simulate_classification_retirements.py

import numpy as np

sim_trial = []
num_sim_trials = 100
for i in range(num_sim_trials): 
    #initiate a dictionary of subject ids with counts all set to 0
    subjects = 3500
    subject_counts = {}
    for i in range(subjects):
        subject_counts[i] = 0
    #Initiate some other parameters
    retired_count = 0
    num_classifications = 0
    limit = 6
    count = 0



   
    while count < limit*subjects:
        x = np.random.randint(0,subjects)
        if subject_counts[x]< limit:
            subject_counts[x] = subject_counts[x]+1
            count = count +1
            if subject_counts[x] == limit:
                retired_count = retired_count + 1
                #print(retired_count," items retired! After ",100*count/(limit*subjects),"% classifications.")
                #if retired_count == 1:
                sim_trial.append(100*count/(limit*subjects))


print(np.mean(sim_trial[1::3500]))
print(np.std(sim_trial[1::3500]))