scikit-learn-whiskers

A collection (only one at this time) of tools aimed to help with some tasks of machine learning and datascience studies.
These tools are intended to be compatible with scikit-learn utilities, and work properly inside a Pipeline.

WhiskerOutliers

A class to mark as outliers the values that can visually be identified as outliers from a typical box and whiskers plot.
This class implements .fit, transform and fit_transform, as well as get_params and set_params methods as any standard scikit-learn implementation.

StandardOutliers

A class to mark as outliers the values outside the range threshold * standard deviation around the mean.
This class implements .fit, transform and fit_transform, as well as get_params and set_params methods as any standard scikit-learn implementation.

Requisites:

• NumPy
• Pandas
• Scikit-Learn

Installation

To install it: pip git+https://github.com/ayaranitram/scikit-learn-whiskers