The Argument Over Parameters

There are only 2 hard things in computer science: cache invalidation, naming things, and off-by-one errors.

Hey pips!

Functions. There’s so much to talk about about functions. Wait. Uh…

can().talk(about = "functions", amount = "so much")

hehehehehe. When the language fails, code will illustrate unambiguously.

So, other than compartmentalisation / abstraction / whatever you wanna call it, the key purpose of functions is reusability. We can define an operation, and then use it as many times as we need to, anywhere. Even more importantly, we can define a function to take in parameters, which allow us to vary its behaviour.

def complete_homework(
    subject,
    date_set,
    speed = "normal",
    distracted = False,
    time_limit = None,
):
    ...
    return True

You can provide default values for parameters by setting them = to one in the definition. You can then omit them when calling the function:

>>> complete_homework("maths", "25-04-02")
True

>>> complete_homework("maths", "25-04-02", distracted = True)
True

If you do, Python will just automatically fallback to the default value.

A really common pitfall with this is using a mutable object as a default value – for instance, a list.

def update(items = []):
    items.append("sup")
    return items

Calling this the first time is fine:

>>> update()
["sup"]

But when we call it again…

>>> update()
["sup", "sup"]

Woah, it’s appended twice. And again?

>>> update()
["sup", "sup", "sup"]

>>> update()
["sup", "sup", "sup", "sup"]

Strange, since we only call .append() once in the function. And yet the .append() seems to carry over between calls.

This is because the default value of [] isn’t created every time the update() is called, but rather when it’s defined. This means every call to the function will access the same list object, and so any modifications to it will be reflected in other function calls too.

The workaround is to set the default value as None and dynamically initialise a new list if necessary:

def updated_update(items = None):
    if items is None:
        items = []

    items.append("yo")
    return items

And now:

>>> updated_update()
["yo"]

>>> updated_update()
["yo"]

Finally, you’ll often hear the words parameter and argument both used to describe what you pass into a function.

For those who care, there is a technical difference. “Parameter” refers to the particular variable being set, while “argument” refers to the actual value it’s set to.

def illustrate(drawing = "blue-corner")
    ...        # param   # argument

As such, we might talk about a “required” or “optional” parameter, but a “missing” or “invalid” argument.

illustrate(
    # setting parameter
    drawing = "oshi-no-ko")
              # passing in argument

Not that important, but pedantics can be important o7

Further Reading