Skip to main content

A library wrapping libffm

Project description

pylibffm

pylibffm is a wrapper around libffm to allow using scipy and numpy arrays as input. pylibffm requires SSE and openmp.

Installing

If you're running python 3.7~3.11 on linux

pip install pylibffm

Otherwise

git clone --recurse-submodules https://github.com/Sinacam/pylibffm
cd pylibffm
make
pip install .

Currently, the distribution is incredibly hacky and cannot be pip installed from source. Windows build support is lacking, but code should be correct.

Documentation

The API consists of

  • train
  • load
  • Model
    • Model.save
    • Model.predict
  • make_bin
  • train_with_bin

Use help or read their docstring for their usage.

Diff with libffm

To be deterministic, set openmp threads to 1. For pylibffm, do

OMP_NUM_THREADS=1 python <script>.py

For libffm, run with -s 1 (the default). This should yield identical models.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pylibffm-0.4.1.tar.gz (543.5 kB view details)

Uploaded Source

File details

Details for the file pylibffm-0.4.1.tar.gz.

File metadata

  • Download URL: pylibffm-0.4.1.tar.gz
  • Upload date:
  • Size: 543.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.8.11

File hashes

Hashes for pylibffm-0.4.1.tar.gz
Algorithm Hash digest
SHA256 d54458ef8b4a292d83f009a0f165b950c92ef744f86e6ac72516c761542012c6
MD5 9a0c19baa35ac2579a60fe368fe8fa65
BLAKE2b-256 b742352cda1d0769eaa993ca498b9dd80b25c72c27da2af33e25683effae67eb

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page