lestpy 0.0.11

Regression model with logical interactions

lestpy

Overview

Why

Lestpy is aimed to model regression problems with a bunch of logical interactions. This approach is an alternative way to describe the relationship between the features and the targets.

Getting Started

Installation

You can install LestPy from PyPI:

    python -m pip install lestpy

Last release

Lespy 0.0.11

    python -m pip install lestpy --upgrade

Main classes, methods and attributes

• Interaction Methods:
  1. compute
  2. get_interaction_list
  3. get_interaction_dict
  4. add_interaction_dict
  5. remove_interactions
• InteractionBuilder
• Transformer Methods:
  1. fit
  2. transform
  3. fit_transform
  4. inverse_transform
• LBM_Regression Methods:
  1. transform
  2. fit
  3. fit_transform
  4. predict
  5. optimize
  6. features_analysis
  7. print_model
  8. fitting_score
  9. extract_features
• Display Methods:
  1. ternary_diagram
  2. response_surface
  3. pareto_frontier
  4. sensibility_analysis
  5. display_interaction
  6. residues
  7. fit
  8. metrics_curve
  9. describe
  10. corr_graph (in development)
• Outliers_Inspection Methods:
  1. cooks_distance
  2. mahalanobis_distance
  3. z_score

How to

Use

Lestpy is designed to be used similarly to sklearn modelization classes and their methods (fit(), transform(), predict(), ...)

Contributing

There are many ways to support the development of lestpy:

• File an issue on Github, if you encounter problems, have a proposal, etc.
• Send an email with ideas to the author.
• Submit a pull request on Github if you improved the code and know how to use git.

Links

The source code and issue tracker of this package is to be found on Github: pyonysos/lestpy.