🧠 Dataclasses: automatic creation of methods and properties

Do you hate writing monotonous __init__, __repr__ and __eq__ for each class? Dataclasses do it for you.

😩 Manual implementation is a boring and stupid task

class Point:
    def __init__(self, x, y):
        self.x = x
        self.y = y
    def __repr__(self):
        return f"Point(x={self.x}, y={self.y})"
    def __eq__(self, other):
        return self.x == other.x and self.y == other.y

class User:
    def __init__(self, name, age):
        self.name = name
        self.age = age
    def __repr__(self):
        return f"User(name={self.name}, age={self.age})"
    def __eq__(self, other):
        return self.name == other.name and self.age == other.age

Problem:
This is crap. Tons of boilerplate code that's easy to break or forget to update.

✔️ Correctly (via @dataclass)

from dataclasses import dataclass

@dataclass
class Point:
    x: int
    y: int

@dataclass
class User:
    name: str
    age: int

p1 = Point(10, 20)
p2 = Point(10, 20)
u = User("Ivan", 30)

print(p1)          # Point(x=10, y=20)
print(p1 == p2)    # True
print(u)           # User(name='Ivan', age=30)

How it works:
The decorator @dataclass automatically generates methods based on type annotations.

Customizing a dataclass:

from dataclasses import dataclass, field

@dataclass(order=True, frozen=True)
class Product:
    name: str
    price: float = 0.0
    tags: list[str] = field(default_factory=list, compare=False)

    def expensive(self):
        return self.price > 1000

p1 = Product("Laptop", 1500.0)
p2 = Product("Mouse", 50.0)
print(p1 > p2)        # True (price comparison due to order=True)
p1.tags.append("tech")
# p1.name = "PC"      # Error! frozen=True makes the object immutable

inits generated by default?:

🔵 __initreprtializer with parameters
🔵 __repr_eqetty string representation
🔵 __eq__ - comparison acltllles
gth geTrue: __lt__, __le__, __gt__, __ge__

Important:
Dataclasses

are not a replacement for regular classes. Use them for data structures where standard methods are needed.

👩‍💻 @DataScience4