AutoImpute

AutoImpute - Missing Data Imputation Framework for Machine Learning

AutoImpute is a comprehensive Python library designed to simplify and automate the process of handling missing data in machine learning datasets. It provides a suite of imputation strategies, from simple statistical methods to advanced predictive modeling, ensuring your data is ready for analysis.

Features

• Diverse Imputation Strategies: Supports mean, median, mode, constant, and predictive imputation (using models like XGBoost, CatBoost, LightGBM).
• Automated Workflow: Streamlines the imputation process, fitting seamlessly into scikit-learn pipelines.
• Evaluation Metrics: Includes tools to evaluate imputation quality and impact on model performance.
• Visualization: Visualize missing data patterns and imputation results.

Installation

pip install autoimpute

Usage

from autoimpute.imputation import AutoImputer
import pandas as pd

# Load your data
data = pd.read_csv("your_data.csv")

# Initialize and fit imputer
imputer = AutoImputer()
imputed_data = imputer.fit_transform(data)

print(imputed_data.head())

License

MIT License