多处理和莳萝可以一起做什么？答案

【问题标题】：What can multiprocessing and dill do together?多处理和莳萝可以一起做什么？
【发布时间】：2013-11-27 20:19:43
【问题描述】：

我想在 Python 中使用 multiprocessing 库。可悲的是multiprocessing 使用了pickle，它不支持带有闭包、lambda 的函数或__main__ 中的函数。这三个对我来说都很重要

In [1]: import pickle

In [2]: pickle.dumps(lambda x: x)
PicklingError: Can't pickle <function <lambda> at 0x23c0e60>: it's not found as __main__.<lambda>

幸运的是，dill 更健壮的泡菜。显然dill 在导入时执行魔术以使泡菜工作

In [3]: import dill

In [4]: pickle.dumps(lambda x: x)
Out[4]: "cdill.dill\n_load_type\np0\n(S'FunctionType'\np1 ...

这非常令人鼓舞，特别是因为我无法访问多处理源代码。可悲的是，我仍然无法让这个非常基本的示例工作

import multiprocessing as mp
import dill

p = mp.Pool(4)
print p.map(lambda x: x**2, range(10))

这是为什么？我错过了什么？ multiprocessing+dill 组合的具体限制是什么？

J.F Sebastian 的临时编辑

mrockli@mrockli-notebook:~/workspace/toolz$ python testmp.py 
    Temporary Edit for J.F Sebastian

mrockli@mrockli-notebook:~/workspace/toolz$ python testmp.py 
Exception in thread Thread-2:
Traceback (most recent call last):
  File "/home/mrockli/Software/anaconda/lib/python2.7/threading.py", line 808, in __bootstrap_inner
    self.run()
  File "/home/mrockli/Software/anaconda/lib/python2.7/threading.py", line 761, in run
    self.__target(*self.__args, **self.__kwargs)
  File "/home/mrockli/Software/anaconda/lib/python2.7/multiprocessing/pool.py", line 342, in _handle_tasks
    put(task)
PicklingError: Can't pickle <type 'function'>: attribute lookup __builtin__.function failed

^C
...lots of junk...

[DEBUG/MainProcess] cleaning up worker 3
[DEBUG/MainProcess] cleaning up worker 2
[DEBUG/MainProcess] cleaning up worker 1
[DEBUG/MainProcess] cleaning up worker 0
[DEBUG/MainProcess] added worker
[DEBUG/MainProcess] added worker
[INFO/PoolWorker-5] child process calling self.run()
[INFO/PoolWorker-6] child process calling self.run()
[DEBUG/MainProcess] added worker
[INFO/PoolWorker-7] child process calling self.run()
[DEBUG/MainProcess] added worker
[INFO/PoolWorker-8] child process calling self.run()Exception in thread Thread-2:
Traceback (most recent call last):
  File "/home/mrockli/Software/anaconda/lib/python2.7/threading.py", line 808, in __bootstrap_inner
    self.run()
  File "/home/mrockli/Software/anaconda/lib/python2.7/threading.py", line 761, in run
    self.__target(*self.__args, **self.__kwargs)
  File "/home/mrockli/Software/anaconda/lib/python2.7/multiprocessing/pool.py", line 342, in _handle_tasks
    put(task)
PicklingError: Can't pickle <type 'function'>: attribute lookup __builtin__.function failed

^C
...lots of junk...

[DEBUG/MainProcess] cleaning up worker 3
[DEBUG/MainProcess] cleaning up worker 2
[DEBUG/MainProcess] cleaning up worker 1
[DEBUG/MainProcess] cleaning up worker 0
[DEBUG/MainProcess] added worker
[DEBUG/MainProcess] added worker
[INFO/PoolWorker-5] child process calling self.run()
[INFO/PoolWorker-6] child process calling self.run()
[DEBUG/MainProcess] added worker
[INFO/PoolWorker-7] child process calling self.run()
[DEBUG/MainProcess] added worker
[INFO/PoolWorker-8] child process calling self.run()

【问题讨论】：

你试过用if __name__ == "__main__":保护池
@J.F.Sebastian 是的，没有变化。明确地说，我在p = mp.Pool(4) 之前和之后都放置了该行，结果没有变化。
1.添加实际代码（使用警卫） 2. 是否有追溯？ 3.启用日志记录：mp.log_to_stderr().setLevel(logging.DEBUG)
先尝试导入 dill。
@J.F.Sebastian 查看带有回溯的编辑

标签： python multiprocessing pickle dill

【解决方案1】：

multiprocessing 在酸洗方面做出了一些错误的选择。不要误会我的意思，它做出了一些不错的选择，使其能够腌制某些类型，以便它们可以在池的地图功能中使用。然而，由于我们有dill 可以进行酸洗，多处理自己的酸洗变得有点限制。实际上，如果multiprocessing 使用pickle 而不是cPickle... 并且还放弃了一些它自己的酸洗覆盖，那么dill 可以接管并为multiprocessing 提供更完整的序列化。

在此之前，multiprocessing 的一个分支称为 pathos（不幸的是，发布版本有点陈旧）消除了上述限制。 Pathos 还添加了多处理没有的一些不错的功能，例如 map 函数中的多参数。 Pathos 即将发布，经过一些温和的更新——主要是转换为 python 3.x。

Python 2.7.5 (default, Sep 30 2013, 20:15:49) 
[GCC 4.2.1 (Apple Inc. build 5566)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import dill
>>> from pathos.multiprocessing import ProcessingPool    
>>> pool = ProcessingPool(nodes=4)
>>> result = pool.map(lambda x: x**2, range(10))
>>> result
[0, 1, 4, 9, 16, 25, 36, 49, 64, 81]

只是为了炫耀一下pathos.multiprocessing 可以做什么......

>>> def busy_add(x,y, delay=0.01):
...     for n in range(x):
...        x += n
...     for n in range(y):
...        y -= n
...     import time
...     time.sleep(delay)
...     return x + y
... 
>>> def busy_squared(x):
...     import time, random
...     time.sleep(2*random.random())
...     return x*x
... 
>>> def squared(x):
...     return x*x
... 
>>> def quad_factory(a=1, b=1, c=0):
...     def quad(x):
...         return a*x**2 + b*x + c
...     return quad
... 
>>> square_plus_one = quad_factory(2,0,1)
>>> 
>>> def test1(pool):
...     print pool
...     print "x: %s\n" % str(x)
...     print pool.map.__name__
...     start = time.time()
...     res = pool.map(squared, x)
...     print "time to results:", time.time() - start
...     print "y: %s\n" % str(res)
...     print pool.imap.__name__
...     start = time.time()
...     res = pool.imap(squared, x)
...     print "time to queue:", time.time() - start
...     start = time.time()
...     res = list(res)
...     print "time to results:", time.time() - start
...     print "y: %s\n" % str(res)
...     print pool.amap.__name__
...     start = time.time()
...     res = pool.amap(squared, x)
...     print "time to queue:", time.time() - start
...     start = time.time()
...     res = res.get()
...     print "time to results:", time.time() - start
...     print "y: %s\n" % str(res)
... 
>>> def test2(pool, items=4, delay=0):
...     _x = range(-items/2,items/2,2)
...     _y = range(len(_x))
...     _d = [delay]*len(_x)
...     print map
...     res1 = map(busy_squared, _x)
...     res2 = map(busy_add, _x, _y, _d)
...     print pool.map
...     _res1 = pool.map(busy_squared, _x)
...     _res2 = pool.map(busy_add, _x, _y, _d)
...     assert _res1 == res1
...     assert _res2 == res2
...     print pool.imap
...     _res1 = pool.imap(busy_squared, _x)
...     _res2 = pool.imap(busy_add, _x, _y, _d)
...     assert list(_res1) == res1
...     assert list(_res2) == res2
...     print pool.amap
...     _res1 = pool.amap(busy_squared, _x)
...     _res2 = pool.amap(busy_add, _x, _y, _d)
...     assert _res1.get() == res1
...     assert _res2.get() == res2
...     print ""
... 
>>> def test3(pool): # test against a function that should fail in pickle
...     print pool
...     print "x: %s\n" % str(x)
...     print pool.map.__name__
...     start = time.time()
...     res = pool.map(square_plus_one, x)
...     print "time to results:", time.time() - start
...     print "y: %s\n" % str(res)
... 
>>> def test4(pool, maxtries, delay):
...     print pool
...     m = pool.amap(busy_add, x, x)
...     tries = 0
...     while not m.ready():
...         time.sleep(delay)
...         tries += 1
...         print "TRY: %s" % tries
...         if tries >= maxtries:
...             print "TIMEOUT"
...             break
...     print m.get()
... 
>>> import time
>>> x = range(18)
>>> delay = 0.01
>>> items = 20
>>> maxtries = 20
>>> from pathos.multiprocessing import ProcessingPool as Pool
>>> pool = Pool(nodes=4)
>>> test1(pool)
<pool ProcessingPool(ncpus=4)>
x: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]

map
time to results: 0.0553691387177
y: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 196, 225, 256, 289]

imap
time to queue: 7.91549682617e-05
time to results: 0.102381229401
y: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 196, 225, 256, 289]

amap
time to queue: 7.08103179932e-05
time to results: 0.0489699840546
y: [0, 1, 4, 9, 16, 25, 36, 49, 64, 81, 100, 121, 144, 169, 196, 225, 256, 289]

>>> test2(pool, items, delay)
<built-in function map>
<bound method ProcessingPool.map of <pool ProcessingPool(ncpus=4)>>
<bound method ProcessingPool.imap of <pool ProcessingPool(ncpus=4)>>
<bound method ProcessingPool.amap of <pool ProcessingPool(ncpus=4)>>

>>> test3(pool)
<pool ProcessingPool(ncpus=4)>
x: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17]

map
time to results: 0.0523059368134
y: [1, 3, 9, 19, 33, 51, 73, 99, 129, 163, 201, 243, 289, 339, 393, 451, 513, 579]

>>> test4(pool, maxtries, delay)
<pool ProcessingPool(ncpus=4)>
TRY: 1
TRY: 2
TRY: 3
TRY: 4
TRY: 5
TRY: 6
TRY: 7
[0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34]

【讨论】：

我很高兴这个应用程序在莳萝开发人员的脑海中。酸洗的脆弱性确实削弱了多处理能力。我怀疑这个问题阻止了许多开发人员对多处理的探索。
从 GitHub 安装最新版本，我使用了pip install git+https://github.com/uqfoundation/pathos
奇怪的是，这在 python 3 中也不起作用，即使它使用pickle 而不是cPickle。引发了相同的PicklingError 异常，因此mp 设法使用原始的_pickle 模块而不是dill，尽管我的import dill as pickle
@max: 这意味着3.x 中的pickle 是cPickle，其中名称现在只是pickle。我可以看到。叹。也许我知道并忘记了。我希望能尽快完成pathos.multiprocessing 的3.x 版本。
@max：我已更新 pathos 以构建和安装 python 3.x，并且即将发布新版本（即 pip install pathos 将按预期工作）。

【解决方案2】：

您可能想尝试使用 multiprocessing_on_dill 库，它是在后端实现 dill 的多处理分支。

例如，您可以运行：

>>> import multiprocessing_on_dill as multiprocessing
>>> with multiprocessing.Pool() as pool:
...     pool.map(lambda x: x**2, range(10))
... 
[0, 1, 4, 9, 16, 25, 36, 49, 64, 81]

【讨论】：

【解决方案3】：

覆盖多处理模块 Pickle 类

import dill, multiprocessing
dill.Pickler.dumps, dill.Pickler.loads = dill.dumps, dill.loads
multiprocessing.reduction.ForkingPickler = dill.Pickler
multiprocessing.reduction.dump = dill.dump
multiprocessing.queues._ForkingPickler = dill.Pickler

【讨论】：

您的答案可以通过额外的支持信息得到改进。请edit 添加更多详细信息，例如引用或文档，以便其他人可以确认您的答案是正确的。你可以找到更多关于如何写好答案的信息in the help center。
这个对我有用，谢谢。我只需要删除 multiprocessing.queues._ForkingPickler = dill.Pickler，因为它显然不再存在。

【解决方案4】：

我知道这个线程很旧，但是，您不一定必须使用 pathos 模块，正如 Mike McKerns 指出的那样。我也觉得 multiprocessing 使用 pickle 而不是 dill 很烦人，所以你可以这样做：

import multiprocessing as mp
import dill
def helperFunction(f, inp, *args, **kwargs):
    import dill # reimport, just in case this is not available on the new processes
    f = dill.loads(f) # converts bytes to (potentially lambda) function
    return f(inp, *args, **kwargs)
def mapStuff(f, inputs, *args, **kwargs):
    pool = mp.Pool(6) # create a 6-worker pool
    f = dill.dumps(f) # converts (potentially lambda) function to bytes
    futures = [pool.apply_async(helperFunction, [f, inp, *args], kwargs) for inp in inputs]
    return [f.get() for f in futures]

然后，你可以这样使用它：

mapStuff(lambda x: x**2, [2, 3]) # returns [4, 9]
mapStuff(lambda x, b: x**2 + b, [2, 3], 1) # returns [5, 10]
mapStuff(lambda x, b: x**2 + b, [2, 3], b=1) # also returns [5, 10]

def f(x):
    return x**2
mapStuff(f, [4, 5]) # returns [16, 25]

它的工作原理基本上是，将 lambda 函数转换为 bytes 对象，将其传递给子进程，并让它重构 lambda 函数。代码中，我刚刚使用dill序列化函数，但如果需要，你也可以序列化参数。

【讨论】：