How to Set Class Attributes from Variable Arguments (Kwargs) in Python

How can you set class attributes from variable arguments (kwargs) in python

You could update the __dict__ attribute (which represents the instance attributes in the form of a dictionary) with the keyword arguments:

class Bar(object):
    def __init__(self, **kwargs):
        self.__dict__.update(kwargs)

then you can:

>>> bar = Bar(a=1, b=2)
>>> bar.a
1

and with something like:

allowed_keys = {'a', 'b', 'c'}
self.__dict__.update((k, v) for k, v in kwargs.items() if k in allowed_keys)

you could filter the keys beforehand (use iteritems instead of items if you’re still using Python 2.x).

Set class attributes with keyword arguments

The type function can take arguments that allow it to create a new class dynamically:

B = type('B', (), d)

b = B()

print(b.x, b.y)

Output:

1 2

The first argument is the name of the generated class. The second is a tuple containing its base classes. In particular, the following two snippets are (roughly) equivalent:

class D(A, B, C):
    pass

and:

D = type('D', (A, B, C), {})

The last argument is a dict mapping names to attributes (both methods and values).

How do I use **kwargs in Python 3 class __init__ function?

General kwargs ideas

When you load variables with self.var = value, it adds it to an internal dictionary that can be accessed with self.__dict__.

class Foo1:

    def __init__(self, **kwargs):
        self.a = kwargs['a']
        self.b = kwargs['b']

foo1 = Foo1(a=1, b=2)
print(foo1.a)  # 1
print(foo1.b)  # 2
print(foo1.__dict__)  # {'a': 1, 'b': 2}

If you want to allow for arbitrary arguments, you can leverage the fact that kwargs is also a dictionary and use the update() function.

class Foo2:

    def __init__(self, **kwargs):
        self.__dict__.update(kwargs)

foo2 = Foo2(some_random_variable=1, whatever_the_user_supplies=2)
print(foo2.some_random_variable)  # 1
print(foo2.whatever_the_user_supplies)  # 2
print(foo2.__dict__)  # {'some_random_variable': 1, 'whatever_the_user_supplies': 2}

This will prevent you from getting an error when you try to store a value that isn't there

class Foo3:

    def __init__(self, **kwargs):
        self.a = kwargs['a']
        self.b = kwargs['b']

foo3 = Foo3(a=1)  # KeyError: 'b'

If you wanted to ensure that variables a or b were set in the class regardless of what the user supplied, you could create class attributes or use kwargs.get()

class Foo4:

    def __init__(self, **kwargs):
        self.a = kwargs.get('a', None)
        self.b = kwargs.get('b', None)

foo4 = Foo4(a=1)
print(foo4.a)  # 1
print(foo4.b)  # None
print(foo4.__dict__)  # {'a': 1, 'b': None}

However, with this method, the variables belong to the class rather than the instance. This is why you see foo5.b return a string, but it's not in foo5.__dict__.

class Foo5:

    a = 'Initial Value for A'
    b = 'Initial Value for B'

    def __init__(self, **kwargs):
        self.__dict__.update(kwargs)

foo5 = Foo5(a=1)
print(foo5.a)  # 1
print(foo5.b)  # Initial Value for B
print(foo5.__dict__)  # {'a': 1}

If you are giving the users the freedom to specify any kwargs they want, you can iterate through the __dict__ in a function.

class Foo6:

    def __init__(self, **kwargs):
        self.__dict__.update(kwargs)

    def do_something(self):
        for k, v in self.__dict__.items():
            print(f"{k} -> {v}")

foo6 = Foo6(some_random_variable=1, whatever_the_user_supplies=2)
foo6.do_something()
# some_random_variable -> 1
# whatever_the_user_supplies -> 2

However, depending on whatever else you have going on in your class, you might end up with a lot more instance attributes than the user supplied. Therefore, it might be good to have the user supply a dictionary as an argument.

class Foo7:

    def __init__(self, user_vars):
        self.user_vars  = user_vars

    def do_something(self):
        for k, v in self.user_vars.items():
            print(f"{k} -> {v}")

foo7 = Foo7({'some_random_variable': 1, 'whatever_the_user_supplies': 2})
foo7.do_something()
# some_random_variable -> 1
# whatever_the_user_supplies -> 2

Addressing your code

With your updated code, I would suggest using the self.__dict__.update(kwargs) method. Then you can either raise an error when you don't encounter variable you're relying on (option1 method) or you can have a default value for the variable incase it's not defined (option2 method)

class MathematicalModel:
    def __init__(self, var1, var2, var3, **kwargs):
        self.var1 = var1
        self.var2 = var2
        self.var3 = var3
        self.__dict__.update(kwargs)  # Store all the extra variables

class MathematicalModelExtended(MathematicalModel):
    def __init__(self, var1, var2, var3, **kwargs):
        super().__init__(var1, var2, var3, **kwargs)

    def option1(self):
        # Trap error if you need var4 to be specified
        if 'var4' not in self.__dict__:
            raise ValueError("Please provide value for var4")

        x = (self.var1 + self.var2 + self.var3) / self.var4
        return x

    def option2(self):
        # Use .get() to provide a default value when the user does not provide it.
        _var4 = self.__dict__.get('var4', 1)

        x = (self.var1 + self.var2 + self.var3) / self.var4
        return x

a = MathematicalModel(1, 2, 3)
b = MathematicalModelExtended(1, 2, 3, var4=4, var5=5, var6=6)
print(b.option1())  # 1.5
print(b.option2())  # 1.5

Granted, if MathematicalModel will never use anything other than var1, var2, and var3, there's no point in passing the kwargs.

class MathematicalModel:
    def __init__(self, var1, var2, var3, **kwargs):
        self.var1 = var1
        self.var2 = var2
        self.var3 = var3

class MathematicalModelExtended(MathematicalModel):
    def __init__(self, var1, var2, var3, **kwargs):
        super().__init__(var1, var2, var3)
        self.__dict__.update(kwargs)

    def option1(self):
        # Trap error if you need var4 to be specified
        if 'var4' not in self.__dict__:
            raise ValueError("Please provide value for var4")

        x = (self.var1 + self.var2 + self.var3) / self.var4
        return x

    def option2(self):
        # Use .get() to provide a default value when the user does not provide it.
        _var4 = self.__dict__.get('var4', 1)

        x = (self.var1 + self.var2 + self.var3) / self.var4
        return x

a = MathematicalModel(1, 2, 3)
b = MathematicalModelExtended(1, 2, 3, var4=4, var5=5, var6=6)
print(b.option1())  # 1.5
print(b.option2())  # 1.5

Trying to use **kwargs to define attributes in a class

kwargs is just a mapping; it doesn't magically create local variables for your function. You need to index the python dictionary with the desired key.

def __init__(self, **kwargs):
    self.health = kwargs['health']
    self.dodge = kwargs['dodge']
    self.damage = kwargs['damage']
    self.critAdd = kwargs['critAdd']

A dataclass simplifies this:

from dataclasses import dataclass

@dataclass
class Character:
    health: int
    dodge: int
    damage: int
    critAdd: float

This generates your original __init__ automatically.

If you need to do additional work in __init__ after adding the dataclass decorator you can define __post_init__ which a dataclass will call after __init__.

How to properly set Python class attributes from init arguments

This is exactly what the dataclasses module is for. You use it like this:

from dataclasses import dataclass

@dataclass
class Foo:
  attr1: str
  attr2: int
  attr3: bool
  # ...etc...

That gives you a class with an implicit __init__ method that takes
care of all the parameter setting for you. You also get things like a
useful __repr__ and __str__ predefined:

>>> myvar = Foo(attr1="somevalue", attr2=42, attr3=False)
>>> myvar
Foo(attr1='somevalue', attr2=42, attr3=False)
>>> print(myvar)
Foo(attr1='somevalue', attr2=42, attr3=False)

Python 2.7: Is it good practice to set all args/kwargs attribute with setattr?

Prior discussion: Python decorator to automatically define __init__ variables, Python: Is it a good idea to dynamically create variables?

Pros:

• Reduces code duplication

Cons:

• Less readable
• Makes it harder to locate member references (thus may also upset static analyzers like pylint)
• The amount of boilerplate code needed to handle all scenarios (e.g. a positional argument passed as a keyword one; check that all required arguments are present) makes code reduction nonexistent and duplicates stock subroutine invocation code
• Requiring a lot of arguments in a constructor that are to be simply set as members is an anti-pattern in and of itself

Class __init__ with mix of fixed and variable length **kwargs arguments

If you pass **kwargs to the constructor it will work, you need to unpack the dict elements into the functions. See also pass **kwargs argument to another function with **kwargs

Proper way to use **kwargs in Python

You can pass a default value to get() for keys that are not in the dictionary:

self.val2 = kwargs.get('val2',"default value")

However, if you plan on using a particular argument with a particular default value, why not use named arguments in the first place?

def __init__(self, val2="default value", **kwargs):

---

Related Topics