Dictionaries: looking things up by name

A list remembers order. A dictionary remembers names.

The thing lists are bad at

Say you are storing one student's marks. With a list you have to remember what each position meant:

marks = [82, 91, 68]
print(marks[1])

91

Ninety-one in what? You have to keep that in your head, or in a comment, and one day you will insert a subject at the front and every position will shift.

A dictionary stores the name alongside the value:

marks = {"maths": 82, "science": 91, "history": 68}

print(marks["science"])
print(marks["history"])

91
68

Curly braces, key: value pairs, commas between. Now the code says what it means, and the order it was written in stops mattering.

Keys, not positions

marks[1] is meaningless here. There is no "item 1" — there is only "science".

marks = {"maths": 82, "science": 91}
print(marks[1])

Read the error: KeyError: 1. Python looked for a key called 1 and did not find one. That is the same shape as list index out of range from the lists lesson: you asked for something that is not there.

Asking for a key that might not exist

This is where dictionaries bite beginners. A missing key stops your program dead:

marks = {"maths": 82}
print(marks["art"])

If you are not certain a key is there, use .get() instead — it hands back None rather than stopping:

marks = {"maths": 82}

print(marks.get("art"))
print(marks.get("art", 0))

None
0

That second argument is a fallback: "give me art, or 0 if there is no art". It is the difference between a program that copes and a program that crashes at three in the morning.

You can also ask directly:

marks = {"maths": 82, "science": 91}

if "art" in marks:
    print("Found art")
else:
    print("No art mark yet")

No art mark yet

Note what in checks: keys, not values. 82 in marks is False, even though 82 is clearly in there.

Adding and changing

Same square brackets, on the left of an =:

marks = {"maths": 82}

marks["art"] = 75          # a key that did not exist — added
marks["maths"] = 88        # a key that did exist — replaced

print(marks)
print(len(marks))

{'maths': 88, 'art': 75}
2

There is no separate "add" and "update". Assigning to a key does whichever one applies, which is convenient but means a typo in a key name adds a new entry instead of raising an error. If a value refuses to change, check your spelling first.

Looping over a dictionary

Loop over a dictionary and you get the keys:

marks = {"maths": 82, "science": 91, "history": 68}

for subject in marks:
    print(subject)

maths
science
history

Usually you want both halves, and .items() gives you them together:

marks = {"maths": 82, "science": 91, "history": 68}

for subject, score in marks.items():
    print(f"{subject}: {score}")

maths: 82
science: 91
history: 68

Two loop variables, because each item is a pair. This is the line you will write most often.

Doing something useful with it

Everything from the loops lesson still applies — the values are just numbers:

marks = {"maths": 82, "science": 91, "history": 68}
total = 0

for subject, score in marks.items():
    if score >= 80:
        print(f"{subject} is strong")
    total += score

print(f"Average: {total / len(marks)}")

maths is strong
science is strong
Average: 80.33333333333333

List or dictionary?

The question to ask is: does position mean anything?

Three test scores where first, second and third are different sittings — that is a list. Three subjects where "second" means nothing — that is a dictionary.

Real programs use both together, and that combination is the next lesson's territory: a list of dictionaries is the shape almost all real data arrives in.