Code Disciples

A blog for all things code

Fri 13 September 2019


Posted by Abhishek Pednekar in Python   

In this post, we will take a look at Python's namedtuple container type. Firstly, namedtuples are immutable containers just like regular tuples. However, unlike regular tuples, which can be accessed only through indices, namedtuples can be accessed via identifiers/keys as well as index values. To use namedtuples, we will need to import the collections module.

Below, we are defining a namedtuple called Movie with four fields in a list - name, genre, year and director. Notice that we are passing Movie as the first argument to the namedtuple method. This parameter is called the typename and is essentially the name of the class being created by calling the namedtuple method.

>>> from collections import namedtuple
>>> Movie = namedtuple("Movie", ["name", "genre", "year", "director"])

Below, is an alternate way to define namedtuples. Here, rather than passing the field names in a list, we are passing them as a string.

>>> from collections import namedtuple
>>> Movie = namedtuple("Movie", "name genre year director")

Using either of the above definitions, we can now create new Movie objects. This is akin to creating a Movie class and providing it with a constructor that accepts values for the four fields.

>>> movie_1 = Movie("The Dark Knight", "Action", 2008, "Christopher Nolan")
>>> movie_2 = Movie("It", "Horror", 2017, "Andy Muschietti")

Accessing namedtuple values

Accessing values using the field identifiers. This is one of the key advantages of namedtuples as it makes the code more readable.

'The Dark Knight'

>>> movie_2.director
'Andy Muschietti'

Accessing values using indices.

>>> movie_1[3]
'Christopher Nolan'

>>> movie_2[0]

Accessing values using getattr().

>>> getattr(movie_1, "year")

>>> getattr(movie_2, "genre")

Built-in methods


The _make() method returns a namedtuple from an iterable. Let's say we have a list containing the attribute values of our Movie instance. By using _make(), we can convert the list to a namedtuple.

>>> movie_lst = ["Taxi Driver", "Thriller", 1976, "Martin Scorsese"]
>>> Movie._make(movie_lst)

Movie(name='Taxi Driver', genre='Thriller', year=1976, director='Martin Scorsese')


The _asdict() method returns the contents of the nametuple as an OrderedDict.

>>> movie_1._asdict()

OrderedDict([('name', 'The Dark Knight'), ('genre', 'Action'), ('year', 2008), 
('director', 'Christopher Nolan')])

The ** (dictionary unpacking) operator

The ** operator can be used to generate a namedtuple from a dictionary. ** unpacks the key-value pairs of a dictionary.

>>> movie_dict = {"name": "March of the Penguins", 
                  "genre": "Documentary", 
                  "year": 2005, 
                  "director": "Luc Jacqet"}

>>> Movie(**movie_dict)

Movie(name='March of the Penguins', genre='Documentary', year=2005, director='Luc Jacqet')


The _fields property is used to return all the field names of a namedtuple.

>>> Movie._fields

('name', 'genre', 'year', 'director')

The _fields property is particularly useful while creating namedtuples that inherit field names from a parent namedtuple. In the example below, we are creating a new namedtuple called NewMovie by inheriting the fields from the original Movie namedtuple using the _fields property. At the same time, we are adding a new field called music to the NewMovie namedtuple.

>>> Movie = namedtuple("Movie", "name genre year director")

>>> NewMovie = namedtuple("NewMovie", Movie._fields + ("music",))

>>> movie_1 = NewMovie("The Dark Knight", "Action", 2008, "Christopher Nolan", "Hans Zimmer")

>>> movie_1
NewMovie(name='The Dark Knight', genre='Action', year=2008, director='Christopher Nolan', music='Hans Zimmer')


The _replace() method is used to selectively replace the value of specific fields.

>>> movie_1._replace(director="C. Nolan")

NewMovie(name='The Dark Knight', genre='Action', year=2008, director='C. Nolan', music='Hans Zimmer')