easyrec recommendations module for django-oscar
Project description
This package provides integration with the recommendation system, easyrec. It is designed to integrate seamlessly with the e-commerce framework django-oscar but can be used without using oscar.
Getting started
Installation
From PyPi:
pip install django-oscar-easyrec
or from Github:
pip install git+git://github.com/tangentlabs/django-oscar-easyrec.git#egg=django-oscar-easyrec
Add 'easyrec' to INSTALLED_APPS
Configuration
Edit your settings.py to set the following settings:
EASYREC_ENDPOINT = 'http://127.0.0.1:8080/easyrec-web/' EASYREC_TENANT_ID = '...' EASYREC_API_KEY = '...'
In easyrec all items have an itemtype. django-oscar-easyrec passes the product class name for this value. If the item type is not registered in easyrec it will send the default value of ‘ITEM’.
So each of your product classes needs to manually added as an itemtype via easyrecs dashboard if you want them to be recorded separately.
Note - if you add item types to easyrec you will need to restart your django project to ensure they are picked up correctly.
And that’s it! All purchases, product views and reviews will automatically be pushed to easyrec.
You can also disable this app by setting EASYREC_HOST to ‘DUMMY’. Useful for testing.
Getting Recommendations
django-oscar-easyrec comes with a templatetag allowing you to easily fetch recommendations and display them in your templates. There are currently 5 supported template tags which do pretty much what they say:
{% load recommendations %}
{% user_recommendations request.user as recommendations %}
{% for recommended_product in recommendations %}
<!-- Do your thing! -->
{% endfor %}
{% users_also_bought a_product request.user as recommendations %}
{% for recommended_product in recommendations %}
<!-- Do your thing! -->
{% endfor %}
{% users_also_viewed a_product request.user as recommendations %}
{% for recommended_product in recommendations %}
<!-- Do your thing! -->
{% endfor %}
{% products_rated_good product as recommendations %}
{% for recommended_product in recommendations %}
<!-- Do your thing! -->
{% endfor %}
{% related_products product as recommendations %}
{% for recommended_product in recommendations %}
<!-- Do your thing! -->
{% endfor %}
Each template tag provides a list of recommended products. If no recommendations are found then an empty QuerySet is returned. Each of these tags also supports a number of other optional parameters.
You can also call the recommendation functions directly:
from easyrec.utils import get_gateway easyrec = get_gateway() recommendations = easyrec.get_user_recommendations(user.user_id) recommendations = easyrec.get_other_users_also_bought(product.upc, user_id) recommendations = easyrec.get_other_users_also_viewed(product.upc, user_id)
user_recommendations
Returns a list of recommended items for a user
parameters:
- user
The user to get recommendations for
- max_results
[optional] The maximum recommendation you wish to receive
- requested_item_type
[optional] The ProductClass of the items you want in the response
- action_type
[optional] The action type you want to get results based on. Valid values are: VIEW, RATE, BUY or any other custom action type you created. Default: VIEW
users_also_bought
Returns a list of recommended items based on users who bought this all bought X
parameters:
- product
The produce you want to find recommendation based on
- user
The request user
- max_results
[optional] The maximum recommendation you wish to receive
- requested_item_type
[optional] The ProductClass of the items you want in the response
users_also_viewed
Returns a list of recommended items based on users who viewed this all viewed X
parameters:
- product
The produce you want to find recommendation based on
- user
The request user
- max_results
[optional] The maximum recommendation you wish to receive
- requested_item_type
[optional] The ProductClass of the items you want in the response
products_rated_good
Returns a list of recommended items based on users who rated this as good also rated X as good.
parameters:
- product
The produce you want to find recommendation based on
- user
The request user
- max_results
[optional] The maximum recommendation you wish to receive
- requested_item_type
[optional] The ProductClass of the items you want in the response
Vagrant
To make testing and development easier I have created a vagrant box with easyrec already installed and configured. If you have vagrant install you can simply perform the following:
vagrant up
The box itself is hosted on Dropbox and the initial download and install will take a long time.
Once the box is up you can acces easyrec using:
http://127.0.0.1:9090/easyrec-web
The username and password to log in are both easyrec. The box also runs MySQL with the password of the root user set to root.
TODO
Dashboard stats
Optional Celery delayed inserts
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Hashes for django-oscar-easyrec-0.0.1.tar.gz
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | 3acac76cd4ec7c2d7742c2e2a7d24d03b82891a15d56b1011d541527950729c2 |
|
| MD5 | dbd29f7f8a2fb72d628f6ccf8031b2ab |
|
| BLAKE2b-256 | 6329f006fcde27e75abf90450ae66d7c80940d1d278d37948fdb3054cd31a909 |
