Last active
March 11, 2024 03:49
-
-
Save JonahTzuChi/12b696a9a14308c81b8a94b977565242 to your computer and use it in GitHub Desktop.
Achieve even distribution in sampling method through mini sorted sampling
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import random as rn | |
from multiprocessing.pool import Pool | |
from queue import Queue | |
import matplotlib.pyplot as plt | |
def sampling(n: int, minIndex: int, maxIndex: int) -> list[int]: | |
lst = [rn.randint(minIndex, maxIndex-1) for _ in range(n)] | |
return lst | |
def extract(main_list: list[int], index_list: list[int]) -> list[tuple[int, int]]: | |
lst = list() | |
try: | |
for i in index_list: | |
p = main_list[i] | |
lst.append((i, p)) | |
return lst | |
except Exception as e: | |
print(f"{e}\n{i}", flush=True) | |
raise e | |
def preprocessing(data: list[int]) -> list[int]: | |
lst = list() | |
for i in range(len(data)): | |
value = data[i] | |
if value == 0: | |
continue | |
for _ in range(value): | |
lst.append(i) | |
return lst | |
def draw_histogram(data, n_bin:int, export_path: str = None): | |
fg, ax = plt.subplots(1, 1) | |
ax.hist(data, bins=50) | |
if export_path: | |
fg.savefig(export_path) | |
def mini_luck(C: int, G: int, T: int, N: int) -> tuple: | |
lst = [0 for i in range(T)] | |
for gen in range(G): | |
try: | |
min_sample = sampling(N, 0, T) | |
sub_list = extract(lst, min_sample) | |
sub_list.sort(key=lambda x: x[1], reverse=False) | |
subject = sub_list[0] | |
subject_index, _ = subject | |
lst[subject_index] += 1 | |
except Exception as e: | |
print(f"[{gen}]\n") | |
raise e | |
x = preprocessing(lst) | |
draw_histogram(x, T, f"./mini_{C}.jpg") | |
SmallestTribeSize = min(lst) | |
LargestTribeSize = max(lst) | |
AverageTribeSize = sum(lst) / TRIBE_SIZE | |
return (SmallestTribeSize, LargestTribeSize, AverageTribeSize) | |
def uniform_luck(C: int, G: int, T: int, N: int = 1): | |
lst = [0 for i in range(T)] | |
for gen in range(G): | |
try: | |
subject_index = int(rn.uniform(0, 1) * (T-1)) | |
lst[subject_index] += 1 | |
except Exception as e: | |
print(f"[{gen}]\n") | |
raise e | |
x = preprocessing(lst) | |
draw_histogram(x, T, f"./uniform_{C}.jpg") | |
SmallestTribeSize = min(lst) | |
LargestTribeSize = max(lst) | |
AverageTribeSize = sum(lst) / TRIBE_SIZE | |
return (SmallestTribeSize, LargestTribeSize, AverageTribeSize) | |
if __name__ == "__main__": | |
N = 10 | |
GENERATION = 10000 | |
TRIBE_SIZE = 3000 | |
CYCLE = 10 | |
PROCESSES = 1 | |
tasks = Queue() | |
with Pool(processes=PROCESSES) as pool: | |
for cycle in range(CYCLE): | |
task = pool.apply_async( | |
uniform_luck, | |
args=( | |
cycle, | |
GENERATION, | |
TRIBE_SIZE, | |
N | |
), | |
callback=lambda x: None, | |
error_callback=lambda x: print(x, flush=True) | |
) | |
tasks.put(task) | |
tasks.put(None) | |
pool.close() | |
pool.join() | |
metrics = [] | |
while True: | |
task = tasks.get() | |
if task is None: | |
break | |
task = task.get() | |
s, l, a = task | |
metrics.append((s, l, a)) | |
SmallestTribeSize = min(list(map(lambda x: x[0], metrics))) | |
LargestTribeSize = max(list(map(lambda x: x[1], metrics))) | |
AverageTribeSize = sum(list(map(lambda x: x[2], metrics))) / CYCLE | |
print(f"SmallestTribeSize: {SmallestTribeSize}") | |
print(f"LargestTribeSize: {LargestTribeSize}") | |
print(f"AverageTribeSize: {AverageTribeSize}") |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Histogram of Mini Luck