Validation

Note that the Python interpreter will not validate the types for you.

However, you can optionally use a type checker like Mypy, which will run static type checks against all your code.

This is similar to running a linter. This can be done locally and ideally as part of your deploy flow.

Benefits of type checking

Mypy can pick up things to alert on you, like:

  • Check if a variable changes type unexpectedly. Like here: 
      x = 'hello'
  x =  2
  • Making sure you don’t pass a bad type e.g. pass None or int type when only str is allowed.
  • Allow a range of types to be allowed (e.g. None or int)
  • Validate types of data structures
  • Create custom types to better represent concepts e.g. A Coordinates type which could be say a tuple of two float values. This makes the code easier to read especially for nested data structure. Note you don’t have to pass in a class instance of Coordinates, you just need to pass in a tuple which matches the structure of the expected type.