aaa-eda-analyzer

Automated exploratory data analysis (EDA) pipeline for pandas DataFrames: data quality checks, outlier detection, bivariate hypothesis testing, mutual information feature importance, multicollinearity (VIF), pairwise feature interactions, intelligent plotting, and a self-contained HTML/PDF report.

📖 Full API reference, guards/thresholds, and detection rules: docs/API.md

Installation

pip install aaa-eda-analyzer

For PDF report export:

pip install "aaa-eda-analyzer[pdf]"

Note: the PyPI/distribution name is aaa-eda-analyzer, but the import name stays eda_analyzer (see Quick start below) — this is normal and matches the pattern used by packages like beautifulsoup4 (imported as bs4).

Quick start

import pandas as pd
from eda_analyzer import DataAnalyzer

df = pd.read_csv("your_data.csv")

analyzer = DataAnalyzer(df, target_col="target")
analyzer.run_all()

Run a subset of steps:

analyzer.run_all(steps=["summarize", "data_quality", "outliers"])

Generate a downloadable HTML report with embedded plots:

analyzer = DataAnalyzer(df, target_col="target", download=True, export_pdf=True)
analyzer.run_all()

Access results programmatically:

analyzer.quality_report_
analyzer.outlier_report_
analyzer.bivariate_stats_
analyzer.feature_importance_
analyzer.vif_report_
analyzer.top_feature_pairs_

License

MIT