discopula

Discrete checkerboard copula modeling and implementation of new scoring methods pertaining to ordinal and categorical discrete data.

Installation

This package (discopula) is hosted on PyPi, so for installation follow the following workflow ...

$ pip install discopula

Now, you should be all set to use it in a Jupyter Notebook!

Alternatively, if you would like to use it in a project, we recommend you to have a virtual environment for your use of this package, then follow the following workflow. For best practices, it's recommended to use a virtual environment:

1. First, create and activate a virtual environment (Python 3.8+ recommended):

# Create virtual environment
$ python -m venv discopula-env

# Activate virtual environment (Mac/Linux)
$ source discopula-env/bin/activate

# Verify you're in the virtual environment
$ which python

2. Install package

$ pip install discopula

3. To deactivate the virtual environment, when done:

$ deactivate

Documentation

Visit Read the Docs for the full documentation, including overviews and several examples.

Examples

For detailed examples in Jupyter Notebooks and beyond (organized by functionality) please refer to our GitHub repository's examples folder.

Features

• Construction of checkerboard copulas from contingency tables and/or list of cases
• Calculation of marginal distributions and CDFs
• Computation of Checkerboard Copula Regression (CCR) and Prediction based on CCR
• Implementation of Checkerboard Copula Regression Association Measure (CCRAM) and the Scaled CCRAM (SCCRAM)
• Bootstrap functionality for CCR-based prediction, CCRAM and SCCRAM
• Permutation testing functionality for CCRAM & SCCRAM
• Vectorized implementations for improved performance
• Rigorous Edge-case Handling & Unit Testing with Pytest

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.