There are two things I missed when I started working with Python after three years of writing Scala code: types and immutability. Fortunately, it turned out that I missed them only because I did not know Python well enough. There is a way of getting some of that functionality in Python without using external libraries!

Table of Contents

Data classes
Functions
Immutability

Spoiler alert. Don’t expect too much. It won’t be like Scala types ;)

Data classes

Let’s start with types. Using data classes, it is possible to specify the type of a field in a class. The most basic usage looks like this:

from dataclasses import dataclass

@dataclass()
class User:
  name: str
  age: int

Seems to be good, but look what happens when I try to assign a string to the age field.

>>> u = User('Test', 'aaa')
>>> u
User(name='Test', age='aaa')

It works! It should not work! There is a more powerful way of defining types which does not work either.

from typing import NewType
Name = NewType('Name', str)
Age = NewType('Age', int)

@dataclass
class User:
  name: Name
  age: Age

>>> User('Name', 123)
User(name='Name', age=123)

>>> User(Name('Test'), Name(1555))
User(name='Test', age=1555)

I am so disappointed.

Functions

At least, we can specify the expected type of a function parameter and the type of the returned value! Can we?

It makes no sense, because still, nothing stops me from misusing it…

def name_to_age(name: Name) -> Age:
  return Age(123)

>>> name_to_age(Age(50))
123

Immutability

Fortunately, there is one thing which works as expected. Python elegantly solves the problem of immutability. All we need to do is adding a parameter to an annotation.

@dataclass(frozen = True)
class User:
  name: str
  age: int

>>> u = User('Test', 123)
>>> u.name = 'Another user'
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<string>", line 3, in __setattr__
dataclasses.FrozenInstanceError: cannot assign to field 'name'

At least something works. What about types? It seems that I can use them only as a part of the documentation. For data validation, https://pydantic-docs.helpmanual.io library must be used.

From Scala to Python - Python dataclasses

Data classes

Functions

Immutability

Notetaking for data science

Re: DataOps Principles: How Startups Do Data The Right Way

From Scala to Python - Python dataclasses

Data classes

Functions

Immutability

Notetaking for data science

Re: DataOps Principles: How Startups Do Data The Right Way

Related Posts

Test-Driven Development in Python with Pytest

Functional programming in Python

Selecting rows in Pandas