Mulitprocess Pools with different functions

Viewed 26528

Most examples of the Multiprocess Worker Pools execute a single function in different processes, f.e.

def foo(args):
   pass

if __name__ == '__main__':
   pool = multiprocessing.Pool(processes=30)
   res=pool.map_async(foo,args)

Is there a way to handle two different and independent functions within the pool? So that you could assign f.e. 15 processes for foo() and 15 processes for bar() or is a pool bounded to a single function? Or du you have to create different processes for different functions manually with

 p = Process(target=foo, args=(whatever,))
 q = Process(target=bar, args=(whatever,))
 q.start()
 p.start()

and forget about the worker pool?

6 Answers

You can use map or some lambda function (edit: actually you can't use a lambda function). You can use a simple map function:

def smap(f, *args):
    return f(*args)

pool = multiprocessing.Pool(processes=30)
res=pool.map(smap, function_list, args_list1, args_list2,...)

The normal map function takes iterables as inputs, which is inconvenient.

Here is a working example of the idea shared by @Rayamon:

import functools

from multiprocessing import Pool


def a(param1, param2, param3):
    return param1 + param2 + param3


def b(param1, param2):
    return param1 + param2


def smap(f):
    return f()


func1 = functools.partial(a, 1, 2, 3)
func2 = functools.partial(b, 1, 2)

pool = Pool(processes=2)
res = pool.map(smap, [func1, func2])
pool.close()
pool.join()
print(res)

Multiple Functions in one Pool

The following example shows how to run the three functions inc, dec, and add in a pool.

from multiprocessing import Pool
import functools

# -------------------------------------

def inc(x):
    return x + 1

def dec(x):
    return x - 1

def add(x, y):
    return x + y

# -------------------------------------

def smap(f):
    return f()

def main():
    f_inc = functools.partial(inc, 4)
    f_dec = functools.partial(dec, 2)
    f_add = functools.partial(add, 3, 4)
    with Pool() as pool:
        res = pool.map(smap, [f_inc, f_dec, f_add])
        print(res)

# -------------------------------------

if __name__ == '__main__':
    main()

We have three functions, which are run independently in a pool. We use the functools.partial to prepare the functions and their parameters before they are executed.

Source: https://zetcode.com/python/multiprocessing/

To further explain the other answer above, here is an example of:

  1. Run a single function with multiple inputs in parallel using a Pool (square function) Interesting Side Note the mangled op on lines for "5 981 25"
  2. Run multiple functions with different inputs (Both args and kwargs) and collect their results using a Pool (pf1, pf2, pf3 functions)
import datetime
import multiprocessing
import time
import random

from multiprocessing import Pool

def square(x):
    # calculate the square of the value of x
    print(x, x*x)
    return x*x

def pf1(*args, **kwargs):
    sleep_time = random.randint(3, 6)
    print("Process : %s\tFunction : %s\tArgs: %s\tsleeping for %d\tTime : %s\n" % (multiprocessing.current_process().name, "pf1", args, sleep_time, datetime.datetime.now()))
    print("Keyword Args from pf1: %s" % kwargs)
    time.sleep(sleep_time)
    print(multiprocessing.current_process().name, "\tpf1 done at %s\n" % datetime.datetime.now())
    return (sum(*args), kwargs)

def pf2(*args):
    sleep_time = random.randint(7, 10)
    print("Process : %s\tFunction : %s\tArgs: %s\tsleeping for %d\tTime : %s\n" % (multiprocessing.current_process().name, "pf2", args, sleep_time, datetime.datetime.now()))
    time.sleep(sleep_time)
    print(multiprocessing.current_process().name, "\tpf2 done at %s\n" % datetime.datetime.now())
    return sum(*args)

def pf3(*args):
    sleep_time = random.randint(0, 3)
    print("Process : %s\tFunction : %s\tArgs: %s\tsleeping for %d\tTime : %s\n" % (multiprocessing.current_process().name, "pf3", args, sleep_time, datetime.datetime.now()))
    time.sleep(sleep_time)
    print(multiprocessing.current_process().name, "\tpf3 done at %s\n" % datetime.datetime.now())
    return sum(*args)

def smap(f, *arg):
    if len(arg) == 2:
        args, kwargs = arg
        return f(list(args), **kwargs)
    elif len(arg) == 1:
        args = arg
        return f(*args)


if __name__ == '__main__':

    # Define the dataset
    dataset = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]

    # Output the dataset
    print ('Dataset: ' + str(dataset))

    # Run this with a pool of 5 agents having a chunksize of 3 until finished
    agents = 5
    chunksize = 3
    with Pool(processes=agents) as pool:
        result = pool.map(square, dataset)
    print("Result of Squares : %s\n\n" % result)
    with Pool(processes=3) as pool:
        result = pool.starmap(smap, [(pf1, [1,2,3], {'a':123, 'b':456}), (pf2, [11,22,33]), (pf3, [111,222,333])])

    # Output the result
    print ('Result: %s ' % result)


Output:
*******

Dataset: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14]
1 1
2 4
3 9
4 16
6 36
7 49
8 64
59 81
 25
10 100
11 121
12 144
13 169
14 196
Result of Squares : [1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 196]


Process : ForkPoolWorker-6  Function : pf1  Args: ([1, 2, 3],)  sleeping for 3  Time : 2020-07-20 00:51:56.477299

Keyword Args from pf1: {'a': 123, 'b': 456}
Process : ForkPoolWorker-7  Function : pf2  Args: ([11, 22, 33],)   sleeping for 8  Time : 2020-07-20 00:51:56.477371

Process : ForkPoolWorker-8  Function : pf3  Args: ([111, 222, 333],)    sleeping for 1  Time : 2020-07-20 00:51:56.477918

ForkPoolWorker-8    pf3 done at 2020-07-20 00:51:57.478808

ForkPoolWorker-6    pf1 done at 2020-07-20 00:51:59.478877

ForkPoolWorker-7    pf2 done at 2020-07-20 00:52:04.478016

Result: [(6, {'a': 123, 'b': 456}), 66, 666] 

Process finished with exit code 0

Related