Contribution and development

YAIB is in active development. The following sections could be relevant for adding new code to our repository

Libraries

The following libraries are important to the operation of YAIB:

• Pandas: Popular data structure framework.
• ReciPys: A modular preprocessing package for Pandas dataframes.
• Pytorch: An open source machine learning framework for deep learning applications.
• Pytorch Lightning: A lightweight Pytorch wrapper for AI research.
• Pytorch Ignite: Library for training and evaluating neural networks in Pytorch.
• Cuda Toolkit: GPU acceleration used for deep learning models.
• Scikit-learn: Machine learning library.
• Scikit-optimize: Used for Bayesian optimization.
• LightGBM: Gradient boosting framework.
• GIN: Provides a lightweight configuration framework for Python.
• Wandb: A tool for visualizing and tracking machine learning experiments.
• Pytest: A testing framework for Python.

Imputation

• HyperImpute: Imputation library for MissForest and GAIN.
• PyPOTS: Imputation library.

Run Tests

python -m pytest ./tests/recipes
coverage run -m pytest ./tests/recipes
# then use either of the following
coverage report
coverage html

Autoformat and lint

For development purposes, we use the ruff package to autoformat our code and as Linting/CI check:

ruff format --line-length 127
ruff check --line-length 127 --statistics