better multiprocessing and multithreading in python


better multiprocessing and multithreading in python

About Multiprocess

multiprocess is a fork of multiprocessing, and is developed as part of pathos:

multiprocessing is a package for the Python language which supports the spawning of processes using the API of the standard library's threading module. multiprocessing has been distributed in the standard library since python 2.6.


  • Objects can be transferred between processes using pipes or multi-producer/multi-consumer queues.

  • Objects can be shared between processes using a server process or (for simple data) shared memory.

  • Equivalents of all the synchronization primitives in threading are available.

  • A Pool class makes it easy to submit tasks to a pool of worker processes.

pathos is a python framework for heterogeneous computing. pathos is in active development, so any user feedback, bug reports, comments, or suggestions are highly appreciated. A list of known issues is maintained at, with a public ticket list at

NOTE: A C compiler is required to build the included extension module. For python 3.3 and above, a C compiler is suggested, but not required.

Major Changes

  • enhanced serialization, using dill

Current Release

This version a fork of multiprocessing-0.70a1.

The latest released pathos fork of multiprocessing is available from::

multiprocessing is distributed under a BSD license.

Development Version

You can get the latest development version with all the shiny new features at::

If you have a new contribution, please submit a pull request.


The multiprocess.Process class follows the API of threading.Thread. For example ::

from multiprocess import Process, Queue

def f(q):
    q.put('hello world')

if __name__ == '__main__':
    q = Queue()
    p = Process(target=f, args=[q])
    print (q.get())

Synchronization primitives like locks, semaphores and conditions are available, for example ::

>>> from multiprocess import Condition
>>> c = Condition()
>>> print (c)
<Condition(<RLock(None, 0)>), 0>
>>> c.acquire()
>>> print (c)
<Condition(<RLock(MainProcess, 1)>), 0>

One can also use a manager to create shared objects either in shared memory or in a server process, for example ::

>>> from multiprocess import Manager
>>> manager = Manager()
>>> l = manager.list(range(10))
>>> l.reverse()
>>> print (l)
[9, 8, 7, 6, 5, 4, 3, 2, 1, 0]
>>> print (repr(l))
<Proxy[list] object at 0x00E1B3B0>

Tasks can be offloaded to a pool of worker processes in various ways, for example ::

>>> from multiprocess import Pool
>>> def f(x): return x*x
>>> p = Pool(4)
>>> result = p.map_async(f, range(10))
>>> print (result.get(timeout=1))
[0, 1, 4, 9, 16, 25, 36, 49, 64, 81]

When dill is installed, serialization is extended to most objects, for example ::

>>> from multiprocess import Pool
>>> p = Pool(4)
>>> print ( x: (lambda y:y**2)(x) + x, xrange(10)))
[0, 2, 6, 12, 20, 30, 42, 56, 72, 90]

More Information

Probably the best way to get started is to look at the examples that are provided within multiprocess. See the examples directory for a set of example scripts. Please feel free to submit a ticket on github, or ask a question on stackoverflow (@Mike McKerns).

pathos is an active research tool. There are a growing number of publications and presentations that discuss real-world examples and new features of pathos in greater detail than presented in the user's guide. If you would like to share how you use pathos in your work, please post a link or send an email (to mmckerns at uqfoundation dot org).


If you use pathos to do research that leads to publication, we ask that you acknowledge use of pathos by citing the following in your publication::

M.M. McKerns, L. Strand, T. Sullivan, A. Fang, M.A.G. Aivazis,
"Building a framework for predictive science", Proceedings of
the 10th Python in Science Conference, 2011;

Michael McKerns and Michael Aivazis,
"pathos: a framework for heterogeneous computing", 2010- ;

Please see or for further information.