Python 嵌套函数变量范围

Question

我有以下功能：

def print_hamming_distance(calls):
    #calls is a dictionary
    samples = calls.keys() 
    with Pool(8) as pool: #Parallel Process
        for dist, sample1, sample2 in pool.imap(multi_proc_hamming_distance, itertools.combinations(samples,2)):
            print( dist, sample1, sample2 )   

def multi_proc_hamming_distance(samples): # specifically created function to use with pool
    return hamming_distance(calls[samples[0]],calls[samples[1]]), samples[0], samples[1]

当我在我的代码中调用它们时，我得到了这个错误：

NameError: name 'calls' is not defined

我的印象是嵌套函数可以访问该函数之外的变量。有人可以向我解释为什么我会收到此错误吗？

我意识到解决方案之一就是将字典作为参数传递给第二个函数，这就是我解决问题的方法，但这增加了运行时间。此外，当我在 jupyter 上运行代码而不包装 print_hamming_distance(calls) 时，它起作用了。

不包装我的意思是这样的：

def multi_proc_hamming_distance(samples): # specifically created function to use with pool
    return hamming_distance(calls[samples[0]],calls[samples[1]]), samples[0], samples[1]


#calls is already defined somewhere
samples = calls.keys() 
with Pool(8) as pool: #Parallel Process
    for dist, sample1, sample2 in pool.imap(multi_proc_hamming_distance, itertools.combinations(samples,2)):
        print( dist, sample1, sample2 )

编辑：完全回溯错误

multiprocessing.pool.RemoteTraceback:
"""
Traceback (most recent call last):
  File             "/home/usr/anaconda3/envs/some_env/lib/python3.5/multiprocessing/pool.py", line 119, in worker
result = (True, func(*args, **kwds))
File "/usr/project/pipeline/project_name/distance.py", line 44, in multi_proc_hamming_distance
return hamming_distance(calls[samples[0]],calls[samples[1]]), samples[0], samples[1]
NameError: name 'calls' is not defined
"""

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/usr/project/pipeline/project_name.py", line 264, in <module>
main()
File "/usr/project/pipeline/project_name.py", line 259, in main
distance(param)
File "/usr/project/pipeline/project_name.py", line 169, in distance
distance = get_distance[param.data_type](calls)
File "/usr/project/pipeline/project_name/distance.py", line 37, in get_param_type_distance
for dist, sample1, sample2 in pool.imap(multi_proc_hamming_distance, itertools.combinations(samples,2)):
File "/home/usr/anaconda3/envs/some_env/lib/python3.5/multiprocessing/pool.py", line 731, in next
raise value

Answer 1

是的，嵌套函数可以访问该函数之外的变量。但是在您的情况下，调用变量未在函数内部定义，它只是一个参数，嵌套函数无法访问。您可以通过添加calls = calls 来纠正该问题，如下所示。

def print_hamming_distance(calls):
    #calls is a dictionary
    calls = calls
    samples = calls.keys() 
    with Pool(8) as pool: #Parallel Process
    for dist, sample1, sample2 in pool.imap(multi_proc_hamming_distance, itertools.combinations(samples,2)):
        print( dist, sample1, sample2 )   

    # nested function
    def multi_proc_hamming_distance(calls,samples): # specifically created function to use with pool
        return hamming_distance(calls[samples[0]],calls[samples[1]]), samples[0], samples[1]