Dictionaries & Sets

Dictionaries — Python’s Most Important Structure

A dict maps keys to values with O(1) lookup. JSON objects, API payloads, config, Django querysets’ values — all dicts:

user = {
    "name": "Digamber",
    "channel": "Coding India",
    "subscribers": 10500,
}

user["name"]               # 'Digamber'
user["city"] = "Mohali"    # add a key
user["subscribers"] += 100 # update
del user["city"]           # remove
"name" in user             # True — checks KEYS

Safe Lookups

user["missing"] raises KeyError. Use get() when a key might not exist:

user.get("city")               # None — no crash
user.get("city", "Unknown")    # 'Unknown' — with a default

setdefault inserts a default only if the key is missing — perfect for grouping:

groups = {}
for word in ["apple", "ant", "bat", "bee"]:
    groups.setdefault(word[0], []).append(word)
# {'a': ['apple', 'ant'], 'b': ['bat', 'bee']}

(Or use collections.defaultdict(list) for the same pattern.)

Iterating Dicts

for key in user:                      # keys by default
    print(key)

for key, value in user.items():      # the usual way
    print(f"{key} = {value}")

list(user.keys())
list(user.values())

Dicts preserve insertion order (guaranteed since Python 3.7).

Merging

defaults = {"theme": "dark", "lang": "en"}
prefs    = {"lang": "hi"}

settings = defaults | prefs        # {'theme': 'dark', 'lang': 'hi'}
# right side wins on conflicts

Sets — Unique, Unordered, Fast

tags = {"python", "django", "python"}   # duplicates collapse
len(tags)                # 2
"python" in tags         # True — O(1), much faster than a list
tags.add("react")
tags.discard("django")

The classic one-liner — dedupe a list:

unique = list(set([1, 2, 2, 3, 3, 3]))   # [1, 2, 3] (order not guaranteed)

Set Algebra

backend  = {"python", "sql", "docker"}
frontend = {"javascript", "css", "docker"}

backend | frontend     # union — everything
backend & frontend     # intersection — {'docker'}
backend - frontend     # difference — {'python', 'sql'}
backend ^ frontend     # symmetric diff — in one but not both

Real use: “which users are in group A but not group B?” — one expression instead of nested loops.