No Duplicates Allowed — Sets

By the end of this lesson you can:

• Build a set with curly braces, and remember the empty-set trap: set(), not {}.
• Add and remove items with .add(), .remove() and .discard().
• Strip duplicates from a list in one line — list(set(my_list)).
• Check membership with in faster than a list lets you.

You ran a poll. Forty students named their favourite hawker dish. The raw responses look like this:

responses = ["nasi lemak", "satay", "nasi lemak", "roti canai",
             "satay", "nasi lemak", "cendol", "satay", "milo"]

How would you answer the question "how many different dishes were mentioned"? You could loop and track seen items, or…

unique = set(responses)
print(unique)
print(len(unique))

{'cendol', 'milo', 'nasi lemak', 'roti canai', 'satay'}
5

Building one

Use curly braces with comma-separated items — looks a bit like a dict but with no :.

vowels  = {"a", "e", "i", "o", "u"}
primes  = {2, 3, 5, 7, 11}

You can also build a set from any iterable using set(...):

from_list = set(["a", "b", "a", "c"])
from_str  = set("hello")
print(from_list)    # → {'a', 'b', 'c'}
print(from_str)     # → {'h', 'e', 'l', 'o'}   (note: only ONE 'l')

maybe = {}              # ← this is a dict!
print(type(maybe))      # → <class 'dict'>

definitely = set()      # ← this is an empty set
print(type(definitely)) # → <class 'set'>

Sets refuse duplicates

Add the same thing twice — the set silently ignores the second one. No error, no warning, no duplicate.

s = {"apple"}
s.add("apple")
s.add("banana")
s.add("apple")
print(s)              # → {'apple', 'banana'}
print(len(s))         # → 2

Adding and removing

.add(x) puts x in. .remove(x) takes it out (or crashes if x isn't there). .discard(x) takes x out if present, and silently does nothing otherwise — the safer choice when you're not sure.

flavours = {"milo", "kopi", "teh"}
flavours.add("horlicks")          # ← in
flavours.remove("kopi")           # ← out (must be there)
flavours.discard("not on menu")   # ← no error, just nothing
print(flavours)
# → {'milo', 'teh', 'horlicks'}

Membership — set's superpower

x in some_set is the same syntax as with a list — but on a set it's much faster. Lists scan one by one. Sets jump straight to the answer.

forbidden = {"banana", "durian", "tempoyak"}
order     = "durian"

if order in forbidden:
    print("Sorry, no", order, "on the bus.")
else:
    print("OK, you can bring", order, "on.")

The shape if x in collection: is one of the most common in Python — and it's a set's best move.

What sets don't have

Sets are unordered — there is no set[0], no .index(), no slicing. If you need order, use a list. If you need labels, use a dict. If you only ever ask "is this in?" or "what unique items?" — set.

Use a LIST   when…  order or duplicates matter
Use a TUPLE  when…  the data is fixed
Use a DICT   when…  you need to look up by label
Use a SET    when…  duplicates are noise, membership is the question

The most common one-liner — strip duplicates

You'll write this loads of times in your life. Worth memorising:

raw = ["a", "b", "a", "c", "b", "a"]
unique = list(set(raw))
print(unique)
# → ['a', 'b', 'c']    (order may vary)

Note: turning a list into a set loses the original order. If you need to preserve order, you'll learn a different trick in PY-L3-21.

The story

You're organising a class party. Three forms come back — sign-ups by name. Some students filled in twice. Some withdrew. You want to print:

1. The unique guest list.
2. How many real guests there are.
3. Whether Aisyah is coming.
4. Whether anyone called Iman is on the list.

Save as rsvp.py:

Code

# rsvp.py — clean the guest list with a set

raw = ["Aisyah", "Wei Jie", "Priya", "Aisyah",
       "Iman", "Aizat", "Priya", "Wei Jie",
       "Iman", "Hafiz"]

guests = set(raw)
print("Guest list :", guests)
print("How many   :", len(guests))

# Step 3
if "Aisyah" in guests:
    print("Aisyah is coming.")
else:
    print("No Aisyah.")

# Step 4
if "Iman" in guests:
    print("An Iman is on the list.")

# A withdrawal
guests.discard("Hafiz")        # uses discard in case he wasn't there
print("After Hafiz cancels:", guests)

Output

Guest list : {'Wei Jie', 'Priya', 'Aizat', 'Iman', 'Aisyah', 'Hafiz'}
How many   : 6
Aisyah is coming.
An Iman is on the list.
After Hafiz cancels: {'Wei Jie', 'Priya', 'Aizat', 'Iman', 'Aisyah'}

What just happened

Ten raw sign-ups became six unique guests — the duplicates dropped out the moment we wrapped the list in set(...). From there every question became one short line — len(), in, discard(). The order of names in the printout isn't the order they signed up; sets don't track order, and you might see them printed in a different sequence on your machine.

Three quick exercises.

word = "programming"
letters = set(word)
print(letters)
print("Unique letters:", len(letters))
# → 8

banned = {"spam", "junk", "scam"}
word = input("Word: ").lower()
if word in banned:
    print("Blocked.")
else:
    print("Allowed.")

scans = ["A123", "B456", "A123", "C789", "B456", "D012", "C789", "A123"]

scans = ["A123", "B456", "A123", "C789", "B456", "D012", "C789", "A123"]
unique = sorted(set(scans))
print(unique)             # → ['A123', 'B456', 'C789', 'D012']
print("Real attendance:", len(unique))

A set stores unique items. Build it with curly braces (but use set() for an empty one), add with .add(), remove with .remove() or the gentler .discard(), and ask "is this here?" with in. Sets give you the one-line de-dupe trick list(set(lst)) and very fast membership checks. They have no order, no labels, and no indexing — pick a list, tuple or dict if you need those.