Brigli 0.1.0

Fast, beautiful EDA and ML statistics for Python developers

brigli

Fast, beautiful EDA and ML statistics for Python developers.

Turn 50 lines of pandas + matplotlib into one line.

import brigli

report = brigli.analyze(df)
report.show()   # beautiful visual summary
print(report)   # clean text summary

---

Install

pip install brigli

Quick start

import pandas as pd
import brigli

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

# Full EDA report — stats + visual plots
report = brigli.analyze(df)
report.show()

# Only want specific columns?
report = brigli.analyze(df, columns=["age", "salary", "city"])

# Just histograms
brigli.plot_distributions(df)

# Save to file instead of opening a window
brigli.plot_distributions(df, save_path="distributions.png")

What you get from analyze(df)

For numeric columns: mean, std, min/max, quartiles, skewness, null count.

For categorical columns: unique count, top-N value frequency, null count.

For the whole dataset: shape, memory usage, total missing % .

Why brigli?

• One line — analyze(df) does what normally takes 50+ lines.
• ML-first — metrics chosen for pre-training EDA, not generic stats.
• Clean aesthetic — plots you can drop straight into reports.
• Production-safe — strict input validation, no eval(), thread-safe.
• Typed — full Python type hints throughout.

Requirements

• Python 3.9+
• pandas ≥ 1.5
• numpy ≥ 1.23
• matplotlib ≥ 3.6
• scipy ≥ 1.9

License

MIT — see LICENSE.