Notice: While Javascript is not essential for this website, your interaction with the content will be limited. Please turn Javascript on for the full experience.

PEP 557 -- Data Classes

Title:Data Classes
Author:Eric V. Smith <eric at>
Type:Standards Track

Notice for Reviewers

This PEP and the initial implementation were drafted in a separate repo: Before commenting in a public forum please at least read the discussion listed at the end of this PEP.


This PEP describes an addition to the standard library called Data Classes. Although they use a very different mechanism, Data Classes can be thought of as "mutable namedtuples with defaults".

A class decorator is provided which inspects a class definition for variables with type annotations as defined in PEP 526, "Syntax for Variable Annotations". In this document, such variables are called fields. Using these fields, the decorator adds generated method definitions to the class to support instance initialization, a repr, and comparisons methods. Such a class is called a Data Class, but there's really nothing special about the class: it is the same class but with the generated methods added.

As an example:

class InventoryItem:
    name: str
    unit_price: float
    quantity_on_hand: int = 0

    def total_cost(self) -> float:
        return self.unit_price * self.quantity_on_hand

The @dataclass decorator will add the equivalent of these methods to the InventoryItem class:

def __init__(self, name: str, unit_price: float, quantity_on_hand: int = 0) -> None: = name
    self.unit_price = unit_price
    self.quantity_on_hand = quantity_on_hand
def __repr__(self):
    return f'InventoryItem(name={!r},unit_price={self.unit_price!r},quantity_on_hand={self.quantity_on_hand!r})'
def __eq__(self, other):
    if other.__class__ is self.__class__:
        return (, self.unit_price, self.quantity_on_hand) == (, other.unit_price, other.quantity_on_hand)
    return NotImplemented
def __ne__(self, other):
    if other.__class__ is self.__class__:
        return (, self.unit_price, self.quantity_on_hand) != (, other.unit_price, other.quantity_on_hand)
    return NotImplemented
def __lt__(self, other):
    if other.__class__ is self.__class__:
        return (, self.unit_price, self.quantity_on_hand) < (, other.unit_price, other.quantity_on_hand)
    return NotImplemented
def __le__(self, other):
    if other.__class__ is self.__class__:
        return (, self.unit_price, self.quantity_on_hand) <= (, other.unit_price, other.quantity_on_hand)
    return NotImplemented
def __gt__(self, other):
    if other.__class__ is self.__class__:
        return (, self.unit_price, self.quantity_on_hand) > (, other.unit_price, other.quantity_on_hand)
    return NotImplemented
def __ge__(self, other):
    if other.__class__ is self.__class__:
        return (, self.unit_price, self.quantity_on_hand) >= (, other.unit_price, other.quantity_on_hand)
    return NotImplemented

Data Classes saves you from writing and maintaining these functions.


There have been numerous attempts to define classes which exist primarily to store values which are accessible by attribute lookup. Some examples include:

  • collection.namedtuple in the standard library.
  • typing.NamedTuple in the standard library.
  • The popular attrs [1] project.
  • Many example online recipes [2], packages [3], and questions [4]. David Beazley used a form of data classes as the motivating example in a PyCon 2013 metaclass talk [5].

So, why is this PEP needed?

With the addition of PEP 526, Python has a concise way to specify the type of class members. This PEP leverages that syntax to provide a simple, unobtrusive way to describe Data Classes. With one exception, the specified attribute type annotation is completely ignored by Data Classes.

No base classes or metaclasses are used by Data Classes. Users of these classes are free to use inheritance and metaclasses without any interference from Data Classes. The decorated classes are truly "normal" Python classes. The Data Class decorator should not interfere with any usage of the class.

Data Classes are not, and are not intended to be, a replacement mechanism for all of the above libraries. But being in the standard library will allow many of the simpler use cases to instead leverage Data Classes. Many of the libraries listed have different feature sets, and will of course continue to exist and prosper.

Where is it not appropriate to use Data Classes?

  • Compatibility with tuples is required.
  • True immutability is required.
  • Type validation beyond that provided by PEPs 484 and 526 is required, or value validation is required.

XXX Motivation for each dataclass() and field() parameter


All of the functions described in this PEP will live in a module named dataclasses.

A function dataclass which is typically used as a class decorator is provided to post-process classes and add generated member functions, described below.

The dataclass decorator examines the class to find field's. A field is defined as any variable identified in __annotations__. That is, a variable that is decorated with a type annotation. With a single exception described below, none of the Data Class machinery examines the type specified in the annotation.

Note that __annotations__ is guaranteed to be an ordered mapping, in class declaration order. The order of the fields in all of the generated methods is the order in which they appear in the class.

The dataclass decorator is typically used with no parameters and no parentheses. However, it also supports the following logical signature:

def dataclass(*, init=True, repr=True, hash=None, cmp=True, frozen=False)

If dataclass is used just as a simple decorator with no parameters, it acts as if it has the default values documented in this signature. That is, these three uses of @dataclass are equivalent:

class C:

class C:

@dataclass(init=True, repr=True, hash=None, cmp=True, frozen=False)
class C:

The parameters to dataclass are:

  • init: If true, a __init__ method will be generated.

  • repr: If true, a __repr__ function will be generated. The generated repr string will have the class name and the name and repr of each field, in the order they are defined in the class. Fields that are marked as being excluded from the repr are not included. For example: InventoryItem(name='widget',unit_price=3.0,quantity_on_hand=10).

  • cmp: If true, __eq__, __ne__, __lt__, __le__, __gt__, and __ge__ methods will be generated. These compare the class as if it were a tuple of its fields, in order. Both instances in the comparison must be of the identical type.

  • hash: Either a bool or None. If None (the default), the __hash__ method is generated according to how cmp and frozen are set.

    If cmp and frozen are both true, Data Classes will generate a __hash__ for you. If cmp is true and frozen is false, __hash__ will be set to None, marking it unhashable (which it is). If cmp is false, __hash__ will be left untouched meaning the __hash__ method of the superclass will be used (if superclass is object, this means it will fall back to id-based hashing).

    Although not recommended, you can force Data Classes to create a __hash__ method with hash=True. This might be the case if your class is logically immutable but can nonetheless be mutated. This is a specialized use case and should be considered carefully.

    See the Python documentation [6] for more information.

  • frozen: If True, assigning to fields will generate an exception. This emulates read-only frozen instances. See the discussion below.

field's may optionally specify a default value, using normal Python syntax:

class C:
    a: int       # 'a' has no default value
    b: int = 0   # assign a default value for 'b'

For common and simple use cases, no other functionality is required. There are, however, some Data Class features that require additional per-field information. To satisfy this need for additional information, you can replace the default field value with a call to the provided field() function. The signature of field() is:

def field(*, default=_MISSING, default_factory=_MISSING, repr=True,
          hash=None, init=True, cmp=True)

The _MISSING value is a sentinel object used to detect if the default and default_factory parameters are provided. Users should never use _MISSING or depend on its value. This sentinel is used because None is a valid value for default.

The parameters to field() are:

  • default: If provided, this will be the default value for this field. This is needed because the field call itself replaces the normal position of the default value.
  • default_factory: If provided, it must be a zero-argument callable that will be called when a default value is needed for this field. Among other purposes, this can be used to specify fields with mutable default values, as discussed below. It is an error to specify both default and default_factory.
  • init: If True, this field is included as a parameter to the generated __init__ function.
  • repr: If True, this field is included in the string returned by the generated __repr__ function.
  • cmp: If True, this field is included in the generated comparison methods (__eq__ et al).
  • hash: This can be a bool or None. If True, this field is included in the generated __hash__ method. If None (the default), use the value of cmp: this would normally be the expected behavior. A field needs to be considered in the hash if it's used for comparisons. Setting this value to anything other than None is discouraged.

Field objects

Field objects describe each defined field. These objects are created internally, and are returned by the fields() module-level method (see below). Users should never instantiate a Field object directly. Its attributes are:

  • name: The name of the field.
  • type: The type of the field.
  • default, default_factory, init, repr, hash, and cmp have the identical meaning as they do in the field() declaration.

post-init processing

The generated __init__ code will call a method named __dataclass_post_init__, if it is defined on the class. It will be called as self.__dataclass_post_init__().

Among other uses, this allows for initializing field values that depend on one or more other fields.

Class variables

The one place where dataclass actually inspects the type of a field is to determine if a field is a class variable. It does this by seeing if the type of the field is given as of type typing.ClassVar. If a field is a ClassVar, it is excluded from consideration as a field and is ignored by the Data Class mechanisms.

Frozen instances

It is not possible to create truly immutable Python objects. However, by passing frozen=True to the @dataclass decorator you can emulate immutability. In that case, Data Classes will add __setattr__ and __delattr__ member functions to the class. These functions will raise a FrozenInstanceError when invoked.

There is a tiny performance penalty when using frozen=True: __init__ cannot use simple assignment to initialize fields, and must use object.__setattr__.

Mutable default values

Python stores the default field values in class attributes. Consider this example, not using Data Classes:

class C:
    x = []
    def __init__(self, x=x):
        self.x = x

assert C().x is C().x
assert C().x is not C([]).x

That is, two instances of class C that do not not specify a value for x when creating a class instance will share the same copy of the list. Because Data Classes just use normal Python class creation, they also share this problem. There is no general way for Data Classes to detect this condition. Instead, Data Classes will raise a TypeError if it detects a default parameter of type list, dict, or set. This is a partial solution, but it does protect against many common errors. See How to support mutable default values in the Discussion section for more details.

Using default factory functions is a way to create new instances of mutable types as default values for fields:

class C:
    x: list = field(default_factory=list)

assert C().x is not C().x


When the Data Class is being created by the @dataclass decorator, it looks through all of the class's base classes in reverse MRO (that is, starting at object) and, for each Data Class that it finds, adds the fields from that base class to an ordered mapping of fields. After all of the base classes, it adds its own fields to the ordered mapping. Because the fields are in insertion order, derived classes override base classes. An example:

class Base:
    x: float = 15.0
    y: int = 0

class C(Base):
    z: int = 10
    x: int = 15

The final list of fields is, in order, x, y, z. The final type of x is int, as specified in class C.

Default factory functions

If a field specifies a default_factory, it is called with zero arguments when a default value for the field is needed. For example, to create a new instance of a list, use:

l: list = field(default_factory=list)

If a field is excluded from __init__ (using init=False) and the field also specifies default_factory, then the default factory function will always be called from the generated __init__ function. This happens because there is no other way to give the field a default value.

Module level helper functions

  • fields(class_or_instance): Returns a list of Field objects that define the fields for this Data Class. Accepts either a Data Class, or an instance of a Data Class.
  • asdict(instance): todo: recursion, class factories, etc.
  • astuple(instance): todo: recursion, class factories, etc.


python-ideas discussion

This discussion started on python-ideas [7] and was moved to a GitHub repo [8] for further discussion. As part of this discussion, we made the decision to use PEP 526 syntax to drive the discovery of fields.

Support for automatically setting __slots__?

At least for the initial release, __slots__ will not be supported. __slots__ needs to be added at class creation time. The Data Class decorator is called after the class is created, so in order to add __slots__ the decorator would have to create a new class, set __slots__, and return it. Because this behavior is somewhat surprising, the initial version of Data Classes will not support automatically setting __slots__. There are a number of workarounds:

  • Manually add __slots__ in the class definition.
  • Write a function (which could be used as a decorator) that inspects the class using fields() and creates a new class with __slots__ set.

For more discussion, see [9].

Should post-init take params?

The post-init function __dataclass_post_init__ takes no parameters. This was deemed to be simpler than trying to find a mechanism to optionally pass a parameter to the __dataclass_post_init__ function.

Why not just use namedtuple

  • Any namedtuple can be compared to any other with the same number of fields. For example: Point3D(2017, 6, 2) == Date(2017, 6, 2). With Data Classes, this would return False.

  • A namedtuple can be compared to a tuple. For example Point2D(1, 10) == (1, 10). With Data Classes, this would return False.

  • Instances are always iterable, which can make it difficult to add fields. If a library defines:

    Time = namedtuple('Time', ['hour', 'minute'])
    def get_time():
        return Time(12, 0)

    Then if a user uses this code as:

    hour, minute = get_time()

    then it would not be possible to add a second field to Time without breaking the user's code.

  • No option for mutable instances.

  • Cannot specify default values.

  • Cannot control which fields are used for __init__, __repr__, etc.

Why not just use typing.NamedTuple

For classes with statically defined fields, it does support similar syntax to Data Classes, using type annotations. This produces a namedtuple, so it shares namedtuple's benefits and some of its downsides.

Why not just use attrs

  • attrs moves faster than could be accommodated if it were moved in to the standard library.
  • attrs supports additional features not being proposed here: validators, converters, metadata, etc. Data Classes makes a tradeoff to achieve simplicity by not implementing these features.

For more discussion, see [10].

Dynamic creation of classes

An earlier version of this PEP and the sample implementation provided a make_class function that dynamically created Data Classes. This functionality was later dropped, although it might be added at a later time as a helper function. The @dataclass decorator does not care how classes are created, so they could be either statically defined or dynamically defined. For this Data Class:

class C:
    x: int
    y: int = field(init=False, default=0)

Here is one way of dynamically creating the same Data Class:

cls_dict = {'__annotations__': OrderedDict(x=int, y=int),
            'y': field(init=False, default=0),
C = dataclass(type('C', (object,), cls_dict))

How to support mutable default values

One proposal was to automatically copy defaults, so that if a literal list [] was a default value, each instance would get a new list. There were undesirable side effects of this decision, so the final decision is to disallow the 3 known built-in mutable types: list, dict, and set. For a complete discussion of this and other options, see [11].


This code exists in a closed source project:

class Application:
    def __init__(self, name, requirements, constraints=None, path='', executable_links=None, executables_dir=()): = name
        self.requirements = requirements
        self.constraints = {} if constraints is None else constraints
        self.path = path
        self.executable_links = [] if executable_links is None else executable_links
        self.executables_dir = executables_dir
        self.additional_items = []

    def __repr__(self):
        return f'Application({!r},{self.requirements!r},{self.constraints!r},{self.path!r},{self.executable_links!r},{self.executables_dir!r},{self.additional_items!r})'

This can be replaced by:

class Application:
    name: Str
    requirements: List
    constraints: List[str] = field(default_factory=list)
    path: Str = ''
    executable_links: List[str] = field(default_factory=list)
    executable_dir: Tuple[str] = ()
    additional_items: List[str] = field(init=False, default_factory=list)

The Data Class version is more declarative, has less code, supports typing, and includes the other generated functions.


The following people provided invaluable input during the development of this PEP and code: Ivan Levkivskyi, Guido van Rossum, Hynek Schlawack, Raymond Hettinger, and Lisa Roach. I thank them for their time and expertise.

A special mention must be made about the attrs project. It was a true inspiration for this PEP, and I respect the design decisions they made.