easy-multip 0.3.0

Easy Python multiprocessing

easy_multip

easy_multip is a small tool designed to quickly allow Python multiprocessing capabilities while greatly simplifying code for easier use.

Current Features

• easy_multip.map(func, arg_group, leave_one_cpu_free=True)
  • map() or list comprehension type functionality that is parallelized using multiprocessing and includes a progress bar
  • Usage is similar to the below constructs:

    list(map(func, arg_group))
        or
    [func(arg) for arg in arg_group]
    

• easy_multip.doloop(func, arg_group, leave_one_cpu_free=True)
  • for loop equivalent that runs a function that returns None
  • Useful in situations like file processing where each operation is expensive and totally independent
  • Allocates jobs evenly among processors and provides a progress bar... of sorts
  • Usage is similar to the below construct:

    for arg in arg_group:
        func(arg)
    

• easy_multip.decorators.use_multip(func, leave_one_cpu_free=True)
  • Decorator providing capability of quickly adding multiprocessing to a function operating on a list
  • ONLY for functions taking a list first arg that returns a list or None
  • DO NOT USE @decorator syntax! Must use the following pattern:

    def _func(list_arg, *args, **kwargs):
        # stuff happens
        return [] (or None)
    func = use_multip(_func)
    

License

MIT