Is the Magic / Dunder Methods lesson free?

Yes. The first 5 lessons on PyRun are free forever — no signup, no card. Magic / Dunder Methods runs real Python in your browser via Pyodide.

How long does the Magic / Dunder Methods lesson take?

About 8 minutes end-to-end. It's a hands-on lesson — you read a paragraph, edit the code, and run it in the same tab.

Do I need to install Python to try the Magic / Dunder Methods lesson?

No install required. PyRun runs real CPython in your browser via Pyodide (Python compiled to WebAssembly). Works on Chrome, Firefox, Safari, and Edge — desktop or mobile.

What level is the Magic / Dunder Methods lesson?

Advanced level. Assumes comfort with classes, functions, and standard library. Not for first-time Python learners.

← All lessons

Advanced·8 min·advanced · oop

Magic / Dunder Methods

You’ll be able to

Customize object representation with __repr__ and __str__
Enable arithmetic with __add__, __mul__, __sub__
Support comparison with __eq__, __lt__, functools.total_ordering
Make objects iterable with __iter__ and __next__

Why this matters

Pandas' DataFrame.__getitem__, SQLAlchemy's operator overloading, and pathlib's / operator are all dunder tricks. Staff-level interviews often probe __eq__/__hash__ contracts and the __init__ vs __new__ split — knowing when to override __new__ matters for singletons, immutables, and metaclass work.

Common pitfalls

Overriding __eq__ without __hash__ — silently makes the class unhashable and breaks set/dict keys.
Writing __repr__ that isn't a valid eval() target — good __repr__ should reconstruct the object for debug logs.
Confusing __str__ (human) with __repr__ (developer) — print() falls back to __repr__ only when __str__ is absent.

Python's protocol-based design: every operator and built-in maps to a "dunder" (double-underscore) method.

The ones you'll use most

Syntax	Method
`len(x)`	`__len__`
`x[i]`	`__getitem__`
`for x in y`	`__iter__` (or `__getitem__` as fallback)
`x + y`	`__add__`
`x == y`	`__eq__`
`str(x)` / print	`__str__` then `__repr__`
`x()`	`__call__`
`with x:`	`__enter__`, `__exit__`

The principle

Implement the dunder, get the syntax for free. Your class plugs straight into all of Python.

Try it

Implement __eq__ on Vec so Vec(1,2) == Vec(1,2).
Add __mul__ to scale a vector by a number.

Practice

3 graded exercises

Write the code, click Check. We’ll tell you exactly what to fix.

Progress0 / 3

Exercise 1
Define a class Bag that wraps a list of items. Implement __len__(self) returning the count and __contains__(self, item) returning whether item is inside. Test: b = Bag([1, 2, 3]); print len(b) (=3), 2 in b (True), 9 in b (False).
Exercise 2
Define Point(x, y) with __eq__ (two Points equal when both x and y match) AND __hash__ (hash the (x, y) tuple). Test: p1 = Point(1, 2); p2 = Point(1, 2); s = {p1, p2} — the set should have length 1. Print len(s).
Exercise 3
Define Vector(x, y) with __add__ that returns a new Vector and __repr__ returning "Vector(x=1, y=2)"-style. Test: Vector(1, 2) + Vector(3, 4) — print the result. Expected repr contains x=4 and y=6.

← PreviousInheritance & super()Next →Decorators

Magic / Dunder Methods

You’ll be able to

Customize object representation with __repr__ and __str__
Enable arithmetic with __add__, __mul__, __sub__
Support comparison with __eq__, __lt__, functools.total_ordering
Make objects iterable with __iter__ and __next__

Why this matters

Common pitfalls

Overriding __eq__ without __hash__ — silently makes the class unhashable and breaks set/dict keys.
Writing __repr__ that isn't a valid eval() target — good __repr__ should reconstruct the object for debug logs.
Confusing __str__ (human) with __repr__ (developer) — print() falls back to __repr__ only when __str__ is absent.

Python's protocol-based design: every operator and built-in maps to a "dunder" (double-underscore) method.

The ones you'll use most

Syntax	Method
`len(x)`	`__len__`
`x[i]`	`__getitem__`
`for x in y`	`__iter__` (or `__getitem__` as fallback)
`x + y`	`__add__`
`x == y`	`__eq__`
`str(x)` / print	`__str__` then `__repr__`
`x()`	`__call__`
`with x:`	`__enter__`, `__exit__`

The principle

Implement the dunder, get the syntax for free. Your class plugs straight into all of Python.

Try it

Implement __eq__ on Vec so Vec(1,2) == Vec(1,2).
Add __mul__ to scale a vector by a number.

Practice

3 graded exercises

Write the code, click Check. We’ll tell you exactly what to fix.

Progress0 / 3

Exercise 1
Define a class Bag that wraps a list of items. Implement __len__(self) returning the count and __contains__(self, item) returning whether item is inside. Test: b = Bag([1, 2, 3]); print len(b) (=3), 2 in b (True), 9 in b (False).
Exercise 2
Define Point(x, y) with __eq__ (two Points equal when both x and y match) AND __hash__ (hash the (x, y) tuple). Test: p1 = Point(1, 2); p2 = Point(1, 2); s = {p1, p2} — the set should have length 1. Print len(s).
Exercise 3
Define Vector(x, y) with __add__ that returns a new Vector and __repr__ returning "Vector(x=1, y=2)"-style. Test: Vector(1, 2) + Vector(3, 4) — print the result. Expected repr contains x=4 and y=6.