5 Python: The Absolute Minimum You Must Know
Python looks like a huge language, but effective daily use rests on about eight ideas — names, core types, truthiness, for/if, comprehensions, imports, and one famous gotcha — all on this page. Master these and every library you meet is just more of the same.
5.1 Variables Are Names, Not Boxes
A Python variable is a name bound to an object, not a box containing a value. Assignment never copies; it points another name at the same object:
>>> a = [1, 2, 3]
>>> b = a # b is a second name for the SAME list
>>> b.append(4)
>>> a
[1, 2, 3, 4]That one model explains half of Python’s surprises. Want an actual copy? Ask for one: b = a.copy(). Numbers and strings never surprise you this way — they’re immutable.
5.2 The Core Types
>>> price = 9.99 # float (42 would be an int)
>>> name = "Ada" # str — immutable
>>> tags = ["new", "sale"] # list — ordered, mutable
>>> point = (3, 4) # tuple — ordered, immutable
>>> user = {"name": "Ada", "id": 7} # dict — key -> value lookup
>>> user["name"]
'Ada'Lists hold sequences you’ll loop over; dicts hold things you look up by key. Those two plus strings do 90% of everyday work. None is the “no value here” object — test with is None.
5.3 f-strings: How You Build Text
Put an f before the quote and drop expressions into {}:
>>> name, score = "Ada", 97.4567
>>> f"{name} scored {score:.1f}%"
'Ada scored 97.5%'The :.1f is a format spec — one decimal place. f-strings replaced every older way of formatting ("%s" % name, .format()); if an AI hands you those, ask for f-strings.
5.4 Truthiness
Every object is truthy or falsy: empty things ("", [], {}, 0, None) are false, non-empty things are true. So idiomatic Python tests the object itself — you’ll rarely see if len(tags) == 0: because if not tags: says the same thing:
>>> tags = []
>>> if not tags:
... print("nothing to show")
nothing to show5.5 for and if: Indentation Is the Syntax
Blocks are defined by indentation — no braces, no end — and a for loop walks any collection directly, no index needed:
>>> for word in ["tea", "coffee"]:
... if len(word) > 3:
... print(word, "is long")
... else:
... print(word, "is short")
tea is short
coffee is longNeed positions too? for i, word in enumerate(words):. Numbers? range(5) gives 0–4.
5.6 Comprehensions: Loops as Expressions
A comprehension builds a new list (or dict) in one readable line — it’s the loop above, folded into the brackets:
>>> nums = [3, 1, 4, 1, 5]
>>> [n * n for n in nums if n > 2]
[9, 16, 25]
>>> {word: len(word) for word in ["tea", "coffee"]}
{'tea': 3, 'coffee': 6}Read it left to right — “n squared, for each n in nums, if n > 2” — and reach for one whenever you’re transforming a collection.
5.7 Importing
>>> import math
>>> math.sqrt(16)
4.0import math brings in the module and you use math.sqrt; from pathlib import Path pulls one name out. Avoid from module import * — nobody can tell where names came from.
5.8 Script vs REPL
The >>> blocks on this page are REPL transcripts — run python3 alone and you get a prompt: type an expression, see its value. It’s your scratchpad. Real programs live in files — python3 report.py runs top to bottom, printing nothing you didn’t print(). The line you’ll see everywhere:
if __name__ == "__main__": # true only when THIS file is the one being run,
main() # not when it's merely imported by another file5.9 The Gotcha: Mutable Default Arguments
Default values are evaluated once, when the function is defined — not on every call. A mutable default is therefore shared between calls:
>>> def add_task(task, tasks=[]):
... tasks.append(task)
... return tasks
>>> add_task("write report")
['write report']
>>> add_task("send email") # a fresh list? No — the SAME list
['write report', 'send email']The standard idiom: default to None (def add_task(task, tasks=None):) and create the list inside — if tasks is None: tasks = [] — so every call gets a fresh one.
5.10 Directing the Machine
An informed prompt names the ideas on this page — the types involved, the shape of the data, the idiom you want — so the AI writes the code you meant instead of guessing.
Vague:
"write python to process my data"Informed:
"I have a list of dicts like {'name': 'Ada', 'score': 97.5}. Give me a dict
comprehension mapping name -> score, skipping entries where score is None,
and an f-string that prints each score to one decimal place."5.11 Spot the Confabulation
An AI assistant explains how to customise settings without touching the original:
DEFAULT_SETTINGS = {"verbose": False, "retries": 3}
def get_settings():
settings = DEFAULT_SETTINGS # take a copy of the defaults
settings["verbose"] = True # customise our copy
return settingsWhat’s wrong?
Assignment doesn’t copy — it binds a second name to the same dict, so this silently corrupts the shared defaults for the whole program: DEFAULT_SETTINGS["verbose"] is now True everywhere. The fix is an explicit DEFAULT_SETTINGS.copy(). Names, not boxes.
5.12 Where to Practice
- futurecoder — free, no-signup course that runs Python in your browser, with a debugger that shows what each name points at.
- Exercism’s Python track — free small exercises with automated tests and human mentoring; ideal once the syntax feels stable.
5.13 Quick Reference
| Idea | The minimum |
|---|---|
| Assignment | binds a name to an object — never copies (b = a.copy() copies) |
| Core types | int, float, str, bool, list, tuple, dict, None |
| f-string | f"{name} scored {score:.1f}%" |
| Truthiness | empty/zero/None are false — write if not tags: |
for |
walks any collection; enumerate for indexes, range(n) for numbers |
| Comprehension | [n*n for n in nums if n > 2] — transform + filter in one line |
| Import | import math → math.sqrt; from pathlib import Path |
| REPL vs script | python3 to experiment; python3 file.py to run a program |
| The gotcha | mutable defaults are shared across calls — default to None instead |
That covers the absolute minimum! You can now read and write idiomatic Python, explain its classic trap, and direct an AI at the rest — everything else is help(thing) away.