LightFM

Build status
Linux
OSX (OpenMP disabled)
Windows (OpenMP disabled)

LightFM is a Python implementation of a number of popular recommendation algorithms for both implicit and explicit feedback, including efficient implementation of BPR and WARP ranking losses. It's easy to use, fast (via multithreaded model estimation), and produces high quality results.

It also makes it possible to incorporate both item and user metadata into the traditional matrix factorization algorithms. It represents each user and item as the sum of the latent representations of their features, thus allowing recommendations to generalise to new items (via item features) and to new users (via user features).

For more details, see the Documentation.

Need help? Contact me via email, Twitter, or Gitter.

Installation

Install from pip:

pip install lightfm

or Conda:

conda install -c conda-forge lightfm

Quickstart

Fitting an implicit feedback model on the MovieLens 100k dataset is very easy:

from lightfm import LightFM
from lightfm.datasets import fetch_movielens
from lightfm.evaluation import precision_at_k

# Load the MovieLens 100k dataset. Only five
# star ratings are treated as positive.
data = fetch_movielens(min_rating=5.0)

# Instantiate and train the model
model = LightFM(loss='warp')
model.fit(data['train'], epochs=30, num_threads=2)

# Evaluate the trained model
test_precision = precision_at_k(model, data['test'], k=5).mean()

Articles and tutorials on using LightFM

1. Learning to Rank Sketchfab Models with LightFM
2. Metadata Embeddings for User and Item Cold-start Recommendations
3. Recommendation Systems - Learn Python for Data Science
4. Using LightFM to Recommend Projects to Consultants

How to cite

Please cite LightFM if it helps your research. You can use the following BibTeX entry:

@inproceedings{DBLP:conf/recsys/Kula15,
  author    = {Maciej Kula},
  editor    = {Toine Bogers and
               Marijn Koolen},
  title     = {Metadata Embeddings for User and Item Cold-start Recommendations},
  booktitle = {Proceedings of the 2nd Workshop on New Trends on Content-Based Recommender
               Systems co-located with 9th {ACM} Conference on Recommender Systems
               (RecSys 2015), Vienna, Austria, September 16-20, 2015.},
  series    = {{CEUR} Workshop Proceedings},
  volume    = {1448},
  pages     = {14--21},
  publisher = {CEUR-WS.org},
  year      = {2015},
  url       = {http://ceur-ws.org/Vol-1448/paper4.pdf},
}

Development

Pull requests are welcome. To install for development:

1. Clone the repository: git clone git@github.com:lyst/lightfm.git
2. Setup a virtual environment: cd lightfm && python3 -m venv venv && source ./venv/bin/activate
3. Install it for development using pip: pip install -e . && pip install -r test-requirements.txt
4. You can run tests by running ./venv/bin/py.test tests.
5. LightFM uses black to enforce code formatting and flake8 for linting, see lint-requirements.txt.
6. [Optional]: You can install pre-commit to locally enfore formatting and linting. Install with:

   pip install pre-commit
   pre-commit install

When making changes to the .pyx extension files, you'll need to run python setup.py cythonize in order to produce the extension .c files before running pip install -e ..