montecarlo.py

import itertools
import random
import math

def map_d(c):
	return math.hypot(random.random(), random.random())

def reduce_d(count_inside, d):
	if d < 1:
		return count_inside + 1

	return count_inside

count = 100000000
d_list = itertools.imap(map_d, xrange(0, count))
count_inside = reduce(reduce_d, d_list )

print 4.0 * count_inside / count

montecarlo_celery.py

import itertools
import random
import math

from celery import Celery
from celery import group

app = Celery('hello', broker='redis://localhost:6379/', backend='redis://localhost')


def map_d(c):
    return math.hypot(random.random(), random.random())


@app.task(name='count_list')
def count_list(count):
    d_list = itertools.imap(map_d, xrange(0, count))
    count_inside = sum(1 for d in d_list if d < 1)
    return count_inside


if __name__ == '__main__':
    count = 1000000
    tasks = 1000
    job = group(count_list.s(count) for i in range(0, tasks)).apply_async()
    count_list = job.get()
    total_count_inside = sum(count_list)

    print 4.0 * total_count_inside / (count * tasks)

montecarlo_futures.py

from concurrent import futures
import itertools
import random
import math

def map_d(c):
    return math.hypot(random.random(), random.random())

def reduce_d(count_inside, d):
	if d < 1:
		return count_inside + 1

	return count_inside

count = 100000
with futures.ProcessPoolExecutor(max_workers=4) as executor:
    d_list = executor.map(map_d, xrange(0, count))
    count_inside = sum(1 for d in d_list if d < 1)

print 4.0 * count_inside / count

requirements.txt

amqp==1.4.4
anyjson==0.3.3
argparse==1.2.1
billiard==3.3.0.16
celery==3.1.10
celery-with-redis==3.0
futures==2.1.6
kombu==3.0.14
pytz==2014.2
redis==2.9.1
wsgiref==0.1.2