"Quantity Field for Django using pint library for automated unit conversions"

pip install django-pint==0.7.1


Build Status codecov PyPI Downloads Python Versions PyPI Version Project Status Wheel Build Code Style Black pre-commit status License: MIT Documentation Status

Django Quantity Field

A Small django field extension allowing you to store quantities in certain units and perform conversions easily. Uses pint behind the scenes. Also contains a form field class and form widget that allows a user to choose alternative units to input data. The cleaned_data will output the value in the base_units defined for the field, eg: you specify you want to store a value in grams but will allow users to input either grams or ounces.

Help wanted

I am currently not working with Django anymore. Therefore the Maintenance of this project is not a priority for me anymore. If there is anybody that could imagine helping out maintaining the project, send me a mail.


Requires django >= 3.2, and python 3.7/3.8/3.9/3.10

Tested with the following combinations:

  • Django 3.2 (Python 3.7, 3.8, 3.9, 3.10)
  • Django 4.0 (Python 3.7, 3.8, 3.9, 3.10)


pip install django-pint

Simple Example

Best way to illustrate is with an example

# app/

from django.db import models
from quantityfield.fields import QuantityField

class HayBale(models.Model):
    weight = QuantityField('tonne')

Quantities are stored as float (Django FloatField) and retrieved like any other field

>> bale = HayBale.objects.create(weight=1.2)
>> bale = HayBale.objects.first()
>> bale.weight
<Quantity(1.2, 'tonne')>
>> bale.weight.magnitude
>> bale.weight.units
<Quantity(1200, 'kilogram')>
<Quantity(2645.55, 'pound')>

If your base unit is atomic (i.e. can be represented by an integer), you may also use IntegerQuantityField and BigIntegerQuantityField.

If you prefer exact units you can use the DecimalQuantityField

You can also pass Quantity objects to be stored in models. These are automatically converted to the units defined for the field ( but can be converted to something else when retrieved of course ).

>> from quantityfield.units import ureg
>> Quantity = ureg.Quantity
>> pounds = Quantity(500 * ureg.pound)
>> bale = HayBale.objects.create(weight=pounds)
>> bale.weight
<Quantity(0.226796, 'tonne')>

Use the inbuilt form field and widget to allow input of quantity values in different units

from quantityfield.fields import QuantityFormField

class HayBaleForm(forms.Form):
    weight = QuantityFormField(base_units='gram', unit_choices=['gram', 'ounce', 'milligram'])

The form will render a float input and a select widget to choose the units. Whenever cleaned_data is presented from the above form the weight field value will be a Quantity with the units set to grams (values are converted from the units input by the user). You also can add the unit_choices directly to the ModelField. It will be propagated correctly.

For comparative lookups, query values will be coerced into the correct units when comparing values, this means that comparing 1 ounce to 1 tonne should yield the correct results.

less_than_a_tonne = HayBale.objects.filter(weight__lt=Quantity(2000 * ureg.pound))

You can also use a custom Pint unit registry in your project

# project/

from pint import UnitRegistry

# django-pint will set the DJANGO_PINT_UNIT_REGISTER automatically
# as application_registry
DJANGO_PINT_UNIT_REGISTER = UnitRegistry('your_units.txt')
DJANGO_PINT_UNIT_REGISTER.define('beer_bootle_weight = 0.8 * kg = beer')

# app/

class HayBale(models.Model):
    # now you can use your custom units in your models
    custom_unit = QuantityField('beer')

Note: As the documentation from pint states quite clearly: For each project there should be only one unit registry. Please note that if you change the unit registry for an already created project with data in a database, you could invalidate your data! So be sure you know what you are doing! Still only adding units should be okay.

Set Up Local Testing

As SQL Lite is not very strict in handling types we use Postgres for testing. This will bring up some possible pitfalls using proper databases. To get the test running please install postgresql on your OS. You need to have psycopg2-binary installed (see tox.ini for further requirements) and a user with the proper permissions set. See for an example on HowTo set it up. Or simply run: sudo -u postgres ./

You can also use you local credentials by creating a tests/ file. See test/ for a description.

The testing is done using pytest or tox.

Local development environment with Docker

To run a local development environment with Docker you need to run the following steps: This is helpful if you have troubles installing postgresql or psycopg2-binary.

  1. git clone your fork
  2. run cp .env.example .env
  3. edit .env file and change it with your credentials ( the postgres host should match the service name in docker-file so you can use "postgres" )
  4. run cp tests/ tests/
  5. run docker-compose up in the root folder, this should build and start 2 containers, one for postgres and the other one python dependencies. Note you have to be in the docker group for this to work.
  6. open a new terminal and run docker-compose exec app bash, this should open a ssh console in the docker container
  7. you can run pytest inside the container to see the result of the tests.

Updating the package

Python and Django major versions have defined EOL. To reduce the maintenance burden and encourage users to use version still receiving security updates. Updating these dependencies have to be done in multiple places:

  • Describing it to end users
  • tox.ini: For local testing
  • setup.cfg: For usage with pip and displaying it in PyPa
  • .travis.yaml: For the CI/CD Definition