/pymltk

Python Machine Learning Toolkit

Primary LanguagePythonApache License 2.0Apache-2.0

TravisCI Status Code Coverage Status Requirements Status Documentation Status

pymltk

Python Machine Learning Toolkit

Description

pymltk is a Python package helping data scientists with their daily work of (pre)processing data and building predictive or other machine learning models. It offers various Python functions which implement common operations done by data scientists during their daily work.

All functions of this package ...

  • ... do one thing and (try to) do it well.
  • ... operate on pandas as well as dask dataframes.
  • ... are fully tested and documented.
  • ... offer a clean and consistent UI.

This package was inspired by mlr, a R package which offers similar functionality with respect to data (pre)processing (but in addition offers a lot more).

Function Overview

  • parse_columns: Parsing features with a specified dtype.
  • parse_missings: Parsing specified values as missing values.
  • merge_levels: Merging levels/values of a feature depending on several criteria.
  • impute_missings: Imputing missing values based on several strategies.
  • remove_constants: Removing features with no/low variability.

Installation

The latest stable release of pymltk is available on pypi and can be installed as usual via

pip install pymltk

In addition, the development version is available in this git repository.

Documentation

A detailed documentation of each function provided by pymltk is available on readthedocs.org.

Contribution

Feature requests and bug reports are very welcome. Please open an issue in the github issue tracker of the respository of this project. Pull requests implementing new functionality are, of course, also welcome. Please open in addition also an issue for those.

License

pymltk is licensed under the Apache License Version 2.0. For details please see the file called LICENSE.

Note

This project has been set up using PyScaffold 2.5.6. For details and usage information on PyScaffold see http://pyscaffold.readthedocs.org/.