A minor error that could happen (at least it happened to me), especially when distributing python scripts/functions that carry out a complete analysis, was to call the function directly at the end of the function .py file. The only things a user needed to modify were the input files and parameters. Doing so when you import you'll get the function running immediately. For proper behavior, you simply need to remove the inside call to the function and reserve it for the real calling file/function/portion of code

虽然你不能在不运行代码的情况下使用import;有一种非常快速的方法可以输入你的变量;通过使用numpy。Savez，它将变量作为numpy数组存储在.npz文件中。然后，您可以使用numpy.load加载变量。

在scipy文档中可以看到完整的描述

请注意，这只适用于变量和变量数组，而不适用于方法等。

因为这就是Python的工作原理——class和def等关键字不是声明。相反，它们是实际执行的实时语句。如果它们没有被执行，您的模块将是空的。

惯用的方法是:

# stuff to run always here such as class/def
def main():
    pass

if __name__ == "__main__":
   # stuff only to run when not called via 'import' here
   main()

但是，它确实需要对导入的模块进行源代码控制。

A minor error that could happen (at least it happened to me), especially when distributing python scripts/functions that carry out a complete analysis, was to call the function directly at the end of the function .py file. The only things a user needed to modify were the input files and parameters. Doing so when you import you'll get the function running immediately. For proper behavior, you simply need to remove the inside call to the function and reserve it for the real calling file/function/portion of code

Another option is to use a binary environment variable, e.g. lets call it 'run_code'. If run_code = 0 (False) structure main.py to bypass the code (but the temporarily bypassed function will still be imported as a module). Later when you are ready to use the imported function (now a module) set the environment variable run_code = 1 (True). Use the os.environ command to set and retrieve the binary variable, but be sure to convert it to an integer when retrieving (or restructure the if statement to read a string value),

在main.py:

import os

#set environment variable to 0 (False):
os.environ['run_code'] = '0'


def binary_module():
    #retrieve environment variable, convert to integer
    run_code_val = int(os.environ['run_code'] )
    
    if run_code_val  == 0:
        print('nope. not doing it.')
    if run_code_val == 1:
        print('executing code...')
        # [do something]

.．.在任何加载main.py的脚本中:

import os,main

main.binary_module() 

输出:不。不做。

# now flip the on switch!
os.environ['run_code'] = '1'
main.binary_module()

输出:执行代码…

*注意:上面的代码假设main.py和任何导入它的脚本存在于同一个目录中。

试着从main.py导入所需的函数?所以,

from main import SomeFunction

可能是你在batch.py中命名的函数与main.py中的函数相同，当你导入main.py时，程序运行main.py函数而不是batch.py函数;执行上述操作可以解决这个问题。我希望。