65 lines
1.9 KiB
Python
65 lines
1.9 KiB
Python
import threading
|
|
import time
|
|
import random
|
|
import numpy as np
|
|
from concurrent.futures import ThreadPoolExecutor
|
|
|
|
def gaussian_determinant(matrix):
|
|
n = len(matrix)
|
|
mat = [row[:] for row in matrix]
|
|
|
|
for i in range(n):
|
|
max_row = max(range(i, n), key=lambda r: abs(mat[r][i]))
|
|
mat[i], mat[max_row] = mat[max_row], mat[i]
|
|
|
|
if mat[i][i] == 0:
|
|
return 0
|
|
|
|
for j in range(i + 1, n):
|
|
factor = mat[j][i] / mat[i][i]
|
|
for k in range(i, n):
|
|
mat[j][k] -= mat[i][k] * factor
|
|
|
|
det = 1
|
|
for i in range(n):
|
|
det *= mat[i][i]
|
|
return det
|
|
|
|
def parallel_determinant(matrix, num_threads=4):
|
|
n = len(matrix)
|
|
result = []
|
|
def worker(start_row, end_row):
|
|
partial_det = 1
|
|
for i in range(start_row, end_row):
|
|
partial_det *= matrix[i][i]
|
|
result.append(partial_det)
|
|
|
|
with ThreadPoolExecutor(max_workers=num_threads) as executor:
|
|
rows_per_thread = n // num_threads
|
|
futures = [executor.submit(worker, i * rows_per_thread, (i + 1) * rows_per_thread) for i in range(num_threads)]
|
|
for future in futures:
|
|
future.result()
|
|
|
|
return sum(result)
|
|
|
|
def generate_matrix(size):
|
|
return [[random.randint(1, 10) for _ in range(size)] for _ in range(size)]
|
|
|
|
matrix_sizes = [100, 300, 500]
|
|
num_threads = 4
|
|
|
|
for size in matrix_sizes:
|
|
print(f"\nБенчмарки для матрицы {size}x{size}:")
|
|
|
|
matrix = generate_matrix(size)
|
|
|
|
start = time.time()
|
|
det_seq = gaussian_determinant(matrix)
|
|
end = time.time()
|
|
print(f"Детерминант (последовательно, метод Гаусса): {det_seq}, время: {end - start:.5f} сек")
|
|
|
|
start = time.time()
|
|
det_par = parallel_determinant(matrix, num_threads=num_threads)
|
|
end = time.time()
|
|
print(f"Детерминант (параллельно): {det_par}, время: {end - start:.5f} сек")
|