datasage-ml 0.1.1

A simple baseline machine learning toolkit for tabular datasets.

DataSage ML

DataSage ML is an open-source Python machine learning toolkit I designed and released to simplify baseline model development for structured datasets. It automates preprocessing, feature encoding, model benchmarking, and model selection, enabling faster experimentation for analysts, learners, and early-stage teams.

pip install datasage-ml

Features

• Load CSV files or pandas DataFrames
• Detect numeric and categorical columns automatically
• Handle missing values
• Encode categorical variables safely
• Train baseline classification models
• Compare model performance
• Select the best model by F1 score
• Generate predictions
• Save the best trained model

Installation for local development

python -m venv .venv
.venv\\Scripts\\activate
pip install -e .\[dev]

For PowerShell, if the extras command gives issues, use:

pip install -e .
pip install pytest build twine

Demo Usage

from datasage\_ml import AutoClassifier

clf = AutoClassifier(target="churn")
clf.fit("data/customer\_churn.csv")

print(clf.leaderboard())

clf.save\_best("best\_model.pkl")

Run the example

python examples\\demo\_classification.py

Run tests

python -m pytest

GitHub push

git init
git add .
git commit -m "Initial commit: add DataSage ML package"
git branch -M main
git remote add origin https://github.com/Jubril-Olasunkanmi/datasage-ml.git
git push -u origin main

Project Structure

datasage-ml/
├── src/
│   └── datasage\_ml/
│       ├── \_\_init\_\_.py
│       └── classifier.py
├── examples/
│   └── demo\_classification.py
├── data/
│   └── customer\_churn.csv
├── tests/
│   └── test\_classifier.py
├── pyproject.toml
├── README.md
├── LICENSE
├── CONTRIBUTING.md
├── requirements.txt
└── .gitignore

Author

Ammar Jubril
Technology Consultant | Data Science | Machine Learning | Financial Engineering

GitHub: https://github.com/Jubril-Olasunkanmi