Simple timeseries data type and field.
Serializes small, simple timeseries to a database with Django. Provides the
Timeseries class for manipulating timeseries, and the
TimeseriesField custom field type for serializing.
TimeseriesField on a model:
from django.db import models from django_simple_timeseries import TimeseriesField class Appliance(models.Model): name = models.CharField(max_length=64) temperature = TimeseriesField( resolution_seconds=60 * 60, max_points=24, help_text='Last 24 hours of temperature data', )
You can then access
Timeseries methods on it:
>>> a = Appliance(name='fridge') >>> a.temperature.add(23.2) >>> a.save() >>> # Wait some time. >>> a.temperature.add(26.5) >>> a.save() >>> print(list(a.iter_points())) [ (datetime(2020, 1, 1, 2, 30, 0, tzinfo=<UTC>), 23.2), (datetime(2020, 1, 2, 2, 30, 0, tzinfo=<UTC>), 26.5), ]
This package supports and is tested against the latest patch versions of:
- Python: 3.7, 3.8, 3.9
- Django: 2.2, 3.0, 3.1
- MariaDB: 10.2, 10.3, 10.4, 10.5
- MySQL: 5.7, 8.0
- Oracle: 12.2+ (only tested against 18.104.22.168 SE)
- PostgreSQL: 9.5, 10, 11, 12
- SQLite: 3.9.0+
All database backends are tested with the latest versions of their drivers. SQLite is also tested on GitHub Actions' latest macOS and Windows virtual environments.
pip install django_simple_timeseries
For Django 3.1 and newer, no additional dependencies are required. For earlier versions, this project requires
pip install django_jsonfield_backport
How it works
Timeseries class implements a simple vector-like timeseries. Timeseries data is always contiguous.
Internally, all timeseries instances have:
datetime.datetimecorresponding to the first data point;
.data, the recorded data points (or y-values); and
.resolution, a timedelta which describes the fixed interval between samples.
Samples are added by calling the
add() method. The
add() method ensures contiguousness with the following policy:
- If fewer than
resolutionseconds have elapsed since the most recent sample, the most recent sample is replaced.
- If more than
resolutionseconds have elapsed since the last sample, the vector is extended by the appropriate number of samples (
time_delta % resolution - 1), each which will be recorded as gaps with the value
- In all cases, the vector is trimmed to no more than
TimeseriesField is implemented as, and extends, a
.from_object() serialize a
Timeseries instance to and from plain python objects, which the custom field type transparently implements.
This module is experimental and hasn't been exhaustively tested. It is not intended for large timeseries. Use at your own risk!
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Hashes for django-simple-timeseries-0.2.0.tar.gz
Hashes for django_simple_timeseries-0.2.0-py3-none-any.whl