Python lists: index, slice, append, pop and sort
Build a Python list, index and slice it, grow and shrink it with append, insert and pop, sort it, and learn what each move costs.
A Python list is an ordered collection you can index, slice, grow, shrink and sort. By the end of this page you will build a list, pick items out of it, change it with append(), insert(), pop() and sort(), know which of those are fast, and avoid the trap where b = a is not a copy.
Create, index and slice a list
A list is written in square brackets, and like a string, each item has a position counted from zero. Negative positions count from the end, len() tells you how many items there are, and a slice such as nums[1:3] runs from 1 up to, but not including, 3. Unlike a string, a list is mutable: nums[0] = 5 changes the item in place. Asking for a position past the end raises IndexError: list index out of range.
Grow and shrink a list
append() adds one item to the end. It is fast, O(1), because Python keeps spare room at the end of the list. insert() puts an item at any position, but every item after it has to shift along, so it costs O(n). pop() with no argument removes and returns the last item, which is O(1). pop(0) takes the first item, and then everything shifts down, so that one is O(n).
nums = [30, 10, 20]
nums.append(40)
nums.insert(0, 5)
print(nums)
print(nums[0], nums[-1])
print(nums[1:3])
last = nums.pop()
print(last, nums)
nums.sort()
print(nums)
print(20 in nums)[5, 30, 10, 20, 40]
5 40
[30, 10]
40 [5, 30, 10, 20]
[5, 10, 20, 30]
TrueLines 2 and 3 grow the list, line 4 shows [5, 30, 10, 20, 40], and lines 5 and 6 print the two ends and the slice. Line 8 pops 40 into last, line 10 sorts the list in place, and line 12 asks whether 20 is in it.
sort, sorted and the in test
sort() orders the list in place and returns None, so nums = nums.sort() leaves you with None instead of a list. If you want a new sorted list and want to keep the original, use the function sorted(). Pass reverse=True to go from largest to smallest, or key=len to sort by length.
The in test checks the items one by one from the left, so it is O(n). If you test membership a lot, a set is faster, and sets are coming in the next lesson.
Why b = a is not a copy
b = a does not copy a list. It gives the same list a second name:
a = [1, 2, 3]
b = a
b.append(4)
print(a)
c = a.copy()
c.append(5)
print(a, c)[1, 2, 3, 4]
[1, 2, 3, 4] [1, 2, 3, 4, 5]Appending to b changed a too, because there is only one list. a.copy(), or the slice a[:], makes a real copy, so appending to c leaves a alone.
What you learned
- A list is an ordered, mutable collection in square brackets.
nums[0]indexes,nums[-1]counts from the end, andnums[1:3]slices.append()andpop()at the end are O(1);insert()andpop(0)shift everything, so they are O(n).sort()changes the list and returnsNone;sorted()returns a new list.b = ais not a copy: usea.copy().
Common questions
What happens if I pop from an empty list? [].pop() raises IndexError: pop from empty list. Check the list first with if nums: when it might be empty.
Is slicing past the end an error? No. With nums = [30, 10, 20], nums[1:100] gives [10, 20] and nums[10:] gives an empty list. Only a single index such as nums[10] raises IndexError.
How do I remove an item by value? Use remove(): nums.remove(10) deletes the first 10 it finds and returns None. If the value is not there it raises ValueError: list.remove(x): x not in list.
Can a list hold different types? Yes. [1, "two", 3.0, [4]] is a valid list, and it can even contain another list.
**Does copy() copy the lists inside a list?** No, it makes a shallow copy. With a = [[1], [2]] and b = a.copy(), appending to b[0] also changes a[0], because both outer lists hold the same inner list.
This lesson follows on from Python strings. Next in the series: tuples and sets.
This article goes with the video Lists. Watch it on YouTube.