Pythonic Exception Handling in a function?

Question

I'm confused about the Pythonic approach to Exception handling in a function. As an example, I have a function calc_mean() which is to calculate the mean of a list of numbers. (I know that the statistics module provides a mean() function, but for learning purposes I'm trying to building my own statistics functions).

One obvious problem that can occur is that the function is passed an empty list. I would like to ask which of the following approaches is most Pythonic:

Do nothing here, let the calling function deal with it

def calc_mean(a_list):
    return sum(a_list)/len(a_list)

Look before you Leap with if-else and return None if list is empty

def calc_mean1(a_list):
    if a_list:
        return sum(a_list)/len(a_list)
    else:
        return None

Look before you Leap with if-else and raise a ValueError if list is empty

def calc_mean2(a_list):
    if a_list:
        return sum(a_list)/len(a_list)
    else:
        raise ValueError("Can't calculate the mean of an empy list")

Easier to ask for Forgiveness with try-except, raise a ValueError if list is empty

def calc_mean3(a_list):
    try:
        return sum(a_list)/len(a_list)
    except ZeroDivisionError:
        raise ValueError("Can't calculate the mean of an empy list")

Easier to ask for Forgiveness with try-except, raise a ValueError with a from if list is empty

def calc_mean4(a_list):
    try:
        return sum(a_list)/len(a_list)
    except ZeroDivisionError:
        raise ValueError("Can't calculate the mean of an empy list") from ZeroDivisionError

My gut feeling is that approaches 4 or 5 are the correct ones, and certainly make it easier to deal with multiple exceptions (e.g. TypeError due to non-numeric elements in the list). However, part of me wonders if approach 1 is preferable as it will calculate and return the mean if it can, or otherwise raise an exception which the calling function should handle.

Pythonic Exception Handling in a function?

0 Answers