Subsection 9.10.2 Passing Lists
On the other hand, if you are passing a mutable object, such as a list, to a function, and the function alters the objectβs state, that state change will be visible globally when the function returns. Take a look at the following example, where a list is passed in to a function.
Try stepping through this in Codelens to see what happens. The state of the list referenced by
lst is altered by
changeit, and since
lst is an alias for
mylst,
mylst is affected by the actions taken by the function.
Look closely at this line:
lst[0] = "Manitoba"
That statement modifies the state of
lst by changing the value in slot 0. Although that line may appear to contradict the statement above that βan assignment to a formal parameter inside a function never affects the argument in the caller,β note that there is a difference between assigning to a
slot of a list, and assigning to the list variable itself. To see that difference, try changing that line to the following:
lst = ["Manitoba", "Bisons"]
Then, run again. This time,
mylist is not altered. To understand why, use CodeLens to step carefully through the code and observe how the assignment to
lst causes it to refer to a separate list.
Take a moment to experiment some more with the
changeit function. Change the body of the function to the following:
lst.append(βManitoba Bisonsβ)
Step through using CodeLens. You should see that
mylst is affected by this change, since the state of the list is altered.
Then, try again with this as the body:
lst = lst + ["Manitoba Bisons"]
Step through using CodeLens. Here, we create a new list using the concatenation operator, and
mylst is not affected by the change.
Understanding the techniques that functions can and cannot use to alter the state of mutable parameters is important. You may want to take some time to study the information on this page more thoroughly and play with the examples until you feel confident about your grasp of the material.
Subsection 9.10.4 Ethics & Data Protection
We most often use lists for storing data, because we often want the flexibility of being able to edit and change information. However, sometimes there is important data (private personal data, health data, salaries, etc.) that needs to be protected from tampering. As a programmer, if you need to share data but you want to ensure the data isnβt tampered with, you can send the data as a tuple. This allows a function to make use of the data, but not change it. Consider this example, which makes use of the accumulator pattern, reading from a tuple, but not modifying it:
The avg_salaries function could take a list or a tuple. By passing in a tuple, the programmer ensures the data is not modified. In this case, where the function is in the same file, itβs obvious that the list isnβt modified inside the function. But often you import modules and use functions that other people have written. As a programmer, you need to ensure that any private data you are responsible for isnβt modified inappropriately, and you can do this by using safe structures like tuples.
Checkpoint 9.10.1.
What is the output of the following code fragment?
def myfun(lst):
lst = [1, 2, 3]
mylist = ['a', 'b']
myfun(mylist)
print(mylist)
Checkpoint 9.10.2.
What is the output of the following code fragment?
def myfun(lst):
del lst[0]
mylist = ['a', 'b']
myfun(mylist)
print(mylist)