Created
September 9, 2019 22:02
-
-
Save saethlin/01867d6dfbd17be8f96ef6629ce1ecbd to your computer and use it in GitHub Desktop.
This file contains hidden or 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 numpy as np | |
import numba | |
import time | |
def add_numpy(a, b): | |
return np.sum(a**2 + b**2) | |
@numba.njit(parallel=True, fastmath=True) | |
def add_numba(a, b): | |
result = 0 | |
for i in numba.prange(len(a)): | |
result += a[i]**2 + b[i]**2 | |
return result | |
@numba.njit(fastmath=True, nogil=True) | |
def histogram(data, bins): | |
binned = np.zeros(bins.size, dtype=np.int64) | |
for d in range(len(data)): | |
for b in range(len(bins)): | |
if data[d] < bins[b]: | |
binned[b] += 1 | |
break | |
return binned | |
def parallel_histogram(data, bins): | |
from concurrent.futures import ThreadPoolExecutor | |
import multiprocessing | |
futures = [] | |
with ThreadPoolExecutor() as executor: | |
chunksize = data.size // multiprocessing.cpu_count() | |
for start in range(0, data.size, chunksize): | |
futures.append(executor.submit(histogram, data[start:start+chunksize], bins)) | |
binned = np.zeros(bins.size, dtype=np.int64) | |
import tqdm | |
for f in futures: | |
binned += f.result() | |
return binned | |
x = np.random.rand(1_000_000) | |
y = np.random.rand(1_000_000) | |
start = time.time() | |
for _ in range(100): | |
add_numpy(x, y) | |
print('numpy', time.time() - start) | |
add_numba(x, y) | |
start = time.time() | |
for _ in range(100): | |
add_numba(x, y) | |
print('numba', time.time() - start) | |
np_version, edges = np.histogram(x, bins=np.linspace(0, 1, 101)) | |
nb_version = histogram(x, np.linspace(0, 1, 101)[1:]) | |
par_nb_version = parallel_histogram(x, np.linspace(0, 1, 101)[1:]) | |
print(nb_version - np_version) | |
print(nb_version - par_nb_version) | |
start = time.time() | |
for _ in range(10): | |
np.histogram(x, bins=np.linspace(0, 1, 101)) | |
print('numpy', time.time() - start) | |
histogram(x, bins=np.linspace(0, 1, 101)[1:]) | |
start = time.time() | |
for _ in range(10): | |
histogram(x, bins=np.linspace(0, 1, 101)[1:]) | |
print('numba', time.time() - start) | |
start = time.time() | |
for _ in range(10): | |
parallel_histogram(x, bins=np.linspace(0, 1, 101)[1:]) | |
print('parallel_numba', time.time() - start) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment