The articles this article is based on are:

Data classes in python let you write less code, and they are supported in python from version 3.7 onward.

Regular classes

We will start with a simple class in Python called perro (dog), with an init method that takes the class arguments as attributes:

  • Nombre: String type.
  • Raza: String type.
  • Edad: Integer type.

class Perro:
    nombre: str
    raza: str
    edad: int

    def __init__(self, nombre, raza, edad):
        self.nombre = nombre
        self.raza = raza
        self.edad = edad

Three pets are instantiated with the Perro class:


mascota1 = Perro('RUDY', 'Terrier Australiano', 2)
mascota2 = Perro('TEDDY', 'bichón frisé', 2)
mascota3 = Perro('TEDDY', 'bichón frisé', 2)

Pet 2 is compared with pet 3:

mascota2 == mascota3

False

Even though both pets have the same data in their attributes, a comparison returns false.

One option is to use a method that compares attribute by attribute.

The contents of mascota2 are printed:


print(mascota2)


<__main__.Perro object at 0x7f9c104e7640>

What is shown is the in-memory id of the Perro instance.

To show the contents, each attribute is accessed:

print(mascota2.nombre,mascota2.raza,mascota2.edad)

TEDDY bichón frisé 2

Data classes

To use dataclass you have to install it via pip:

pip install dataclasses

We will take the same Perro class, this time using a dataclass decorator.

Import dataclass:


from dataclasses import dataclass

The Perro class is defined with the same attributes as the previous class:


@dataclass
class Perro:
    nombre: str
    raza: str
    edad: int

The class is defined and each attribute only gets its type, nothing more.

Now the pets are created:


mascota1 = Perro('RUDY', 'Terrier Australiano', 2)
mascota2 = Perro('TEDDY', 'bichón frisé', 2)
mascota3 = Perro('TEDDY', 'bichón frisé', 2)

Now, when comparing pet 2 and 3, it will not return false:


print(mascota2 == mascota3)

True

Next, the contents of mascota2 are printed:


print(mascota2)

Perro(nombre='TEDDY', raza='bichón frisé', edad=2)

Now it does not return the Id; it returns the perro class with the attribute values that were assigned to mascota2.

You can also access the attribute data:

print(mascota2.nombre,mascota2.raza,mascota2.edad)

TEDDY bichón frisé 2

You can add the repr method to the class to give the information as a formatted string:

@dataclass
class Perro:
    nombre: str
    raza: str
    edad: int

    def __repr__(self):
        return f"{self.nombre} {self.raza} ({self.edad})"


mascota1 = Perro('RUDY', 'Terrier Australiano', 2)
mascota2 = Perro('TEDDY', 'bichón frisé', 2)
mascota3 = Perro('TEDDY', 'bichón frisé', 2)


print(mascota2)

TEDDY bichón frisé (2)

Besides comparing whether they are equal, you can compare whether pet 1 is greater than pet 2.


print(mascota1  > mascota2)

This returns an error; this comparison is not supported:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
/tmp/ipykernel_603591/468771088.py in <module>
----> 1 mascota1  > mascota2

TypeError: '>' not supported between instances of 'Perro' and 'Perro'

To be able to compare them, the post_init method is needed. The attribute used for the comparison will be the age; the code will look like this:


from dataclasses import dataclass, field

@dataclass
class Perro:
    sort_index: int = field(init=False)
    nombre: str
    raza: str
    edad: int

    def __repr__(self):
        return f"{self.sort_index} {self.nombre} {self.raza} ({self.edad})"  

    def __post_init__(self):
      self.sort_index = self.edad

The arguments are shown:


mascota1 = Perro('RUDY', 'Terrier Australiano', 3)
mascota2 = Perro('TEDDY', 'bichón frisé', 2)
print(mascota2)

#2 TEDDY bichón frisé (2)


print(mascota2.sort_index,mascota2.nombre,mascota2.raza,mascota2.edad)

#2 TEDDY bichón frisé 2

Let's try comparing again:

print(mascota1 > mascota2)

It returns an error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
/tmp/ipykernel_603591/4142115589.py in <module>
----> 1 mascota1 > mascota2

TypeError: '>' not supported between instances of 'Perro' and 'Perro'

Read-only classes

You can create read-only classes to prevent their initial values from being modified.

To fix this error, the order argument is used in the decorator, and read-only is defined:


@dataclass(order=True, frozen=True)
class Perro:
    sort_index: int = field(init=False)
    nombre: str
    raza: str
    edad: int

    def __post_init__(self):
      object.__setattr__(self, 'sort_index', self.edad)

mascota1 = Perro('RUDY', 'Terrier Australiano', 3)
mascota2 = Perro('TEDDY', 'bichón frisé', 2)

Now the comparison is made and it does not return an error:

print(mascota1  > mascota2)

True

The contents of pet 2 are shown:


print(mascota2)


Perro(sort_index=2, nombre='TEDDY', raza='bichón frisé', edad=2)

If you try to assign a value to pet 2's age, it will return an error:


mascota2.edad = 4


---------------------------------------------------------------------------
FrozenInstanceError                       Traceback (most recent call last)
/tmp/ipykernel_603591/927787478.py in <module>
----> 1 mascota2.edad = 4

<string> in __setattr__(self, name, value)

FrozenInstanceError: cannot assign to field 'edad'

As you can see, dataclasses can make it easier to validate the data in the attributes of the objects you define. The next article will cover the pydantic module, which extends this data validation feature.

Make a donation! If you liked the article you can make a donation with Bitcoin (BTC) using the digital wallet of your choice to the following address: 17MtNybhdkA9GV3UNS6BTwPcuhjXoPrSzV

Or by scanning the QR code from your wallet:

17MtNybhdkA9GV3UNS6BTwPcuhjXoPrSzV