Python multiprocess pool模块报错pickling error问题解决方法分析_Python

Python multiprocess pool模块报错pickling error问题解决方法分析

2021-06-07 00:55Arkenstone Python

这篇文章主要介绍了Python multiprocess pool模块报错pickling error问题解决方法,结合实例形式分析了multiprocess pool模块报错pickling error的原因与解决方法,需要的朋友可以参考下

本文实例讲述了Python multiprocess pool模块报错pickling error问题解决方法。分享给大家供大家参考，具体如下：

问题

之前在调用class内的函数用multiprocessing模块的pool函数进行多线程处理的时候报了以下下错误信息:

PicklingError: Can't pickle <type 'function'>: attribute lookup __builtin__.function failed

查了下官方文档发现python默认只能pickle以下的类型:

None, True, and False
integers, floating point numbers, complex numbers
strings, bytes, bytearrays
tuples, lists, sets, and dictionaries containing only picklable objects
functions defined at the top level of a module (using def, not lambda)
built-in functions defined at the top level of a module
classes that are defined at the top level of a module
instances of such classes whose dict or the result of calling getstate() is picklable (see section -
Pickling Class Instances for details).

函数只能pickle在顶层定义的函数，很明显的class内的函数无法被pickle因此会报错。

				?

									import multiprocessing

									def work():  # top-level 函数

									  print "work!"

									class Foo():

									  def work(self): # 非top-level函数

									    print "work"

									pool1 = multiprocessing.Pool(processes=4)

									foo = Foo()

									pool1.apply_async(foo.work)

									pool1.close()

									pool1.join()

									# 此时报错

									pool2 = multiprocessing.Pool(processes=4)

									pool2.apply_async(work)

									pool2.close()

									pool2.join()

									# 此时工作正常

解决方案

调用pathos包下的multiprocessing模块代替原生的multiprocessing。pathos中multiprocessing是用dill包改写过的，dill包可以将几乎所有python的类型都serialize，因此都可以被pickle。或者也可以自己用dill写一个（有点重复造轮子之嫌啊）

参考

1. https://stackoverflow.com/questions/8804830/python-multiprocessing-picklingerror-cant-pickle-type-function
2. https://docs.python.org/3/library/pickle.html#what-can-be-pickled-and-unpickled
3. https://github.com/uqfoundation/pathos

希望本文所述对大家Python程序设计有所帮助。

原文链接：https://www.cnblogs.com/arkenstone/p/7901129.html