Skip to main content

A computation graph library for feature engineering with Numpy data

Project description

Megatron is a framework for building computation graphs for
feature engineering in machine learning, with Numpy arrays as the data type.
Use Megatron if you want to:
- Do feature engineering in a modular and functional way, building up features one step at a time
- Use disk space to save time by caching feature sets for easy reloading
- Train feature engineering modules on training data and apply them to testing data
- Write custom functions for complex transformations, but access built-in functions for quick and common transformations
- Build feature engineering like you build Keras models (the API is heavily inspired by Keras)
Or any combination of these.
Megatron is distributed under the MIT license.

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

Megatron-0.3.4a0.tar.gz (20.4 kB view details)

Uploaded Source

File details

Details for the file Megatron-0.3.4a0.tar.gz.

File metadata

  • Download URL: Megatron-0.3.4a0.tar.gz
  • Upload date:
  • Size: 20.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/1.11.0 pkginfo/1.4.2 requests/2.19.1 setuptools/40.4.3 requests-toolbelt/0.8.0 tqdm/4.25.0 CPython/3.7.0

File hashes

Hashes for Megatron-0.3.4a0.tar.gz
Algorithm Hash digest
SHA256 4c40a167b5454b4872802ca3cb73dea746cc80f7318a20d684fc029c71ab29b1
MD5 1b8a642cfa6c5dc46ddc9f4493647e4c
BLAKE2b-256 a83988b8192fe18b4cdb539ed17cd87c3ae5482f53e5a4756f8d430bb3813153

See more details on using hashes here.

Supported by

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