Implement a Safe Average Function in Python

Write a Python function `average(table)` that returns the arithmetic mean of a list of numbers and returns `0` when the list is empty. Explain the function's behavior, complexity, and any input assumptions you would clarify before using it in a data pipeline. ### Constraints & Assumptions - The required empty-list result is exactly `0`. - A direct implementation with Python's built-in numeric operations is sufficient. - Do not silently invent missing-value semantics; explain how `None` or `NaN` would be handled if they can appear. ### Clarifying Questions to Ask - Are integers and floating-point values both valid? - Can the input contain `None`, `NaN`, nonnumeric values, or an iterator rather than a list? - Should the function return an integer or floating-point zero for an empty input? - Is numerical overflow or precision important for the expected data size? ### What a Strong Answer Covers - A correct empty-input branch and correct arithmetic mean for nonempty input. - Time complexity proportional to the number of values and constant auxiliary space for a list input. - Explicit behavior for invalid or missing values rather than accidental coercion. - Awareness that production data work may use library functions with different missing-value rules. ### Follow-up Questions 1. How would you adapt the function to ignore `None` values? 2. How would you compute the mean of a stream that cannot be stored in memory? 3. What happens when the input contains `float('nan')`? 4. How would you test this function?