Skip to content

Releases: TyMill/EcoStackML

initial release for Zenodo

08 Apr 09:54
@TyMill TyMill
v0.2.0
fcf0868
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Choose a tag to compare

View all tags
initial release for Zenodo Latest
Latest

🌿 EcoStackML v1.0.0 — Official Release

Author: Tymoteusz Miller
Date: April 2025
License: MIT
Repository: GitHub - EcoStackML

🚀 About This Release

We are proud to announce the first stable release of EcoStackML, a Python-based machine learning framework focused on modular stacking, environmental intelligence, and explainable AI.

EcoStackML bridges data science best practices with real-world problems — from predicting water quality to detecting patterns in complex structured data. It is built with flexibility, reproducibility, and clarity in mind.

This release reflects over a year of research, engineering, and iteration — shaped by hands-on experience in academia, industry, and environmental science.

🔍 What is EcoStackML?

EcoStackML is a stacked machine learning framework designed to simplify the full ML pipeline for:

  • Environmental monitoring and sustainability research
  • Applied data science (finance, marketing, IoT, geospatial)
  • Academic and student projects needing high-quality ML pipelines
  • Any team looking to rapidly prototype models with transparency and performance

🧠 Why use it?

  • Designed for researchers and ML practitioners
  • Combines power of multiple models into robust stackers
  • Provides end-to-end pipelines — from preprocessing to explainability
  • Makes model building more interpretable and reproducible
  • Documented, tested, and easy to extend

✅ Highlights

  • 🔁 Stacked Model Trainer: Random Forest, XGBoost, SVM + Logistic/GBM meta-learners
  • 🧽 Cleaner: Handles missing values, outliers (IQR, Isolation Forest), scaling
  • 🗃 DataLoader: Supports CSV, JSON, Parquet, Hive
  • 🧪 Evaluator: ROC, PR Curve, confusion matrix, metrics
  • 🔍 SHAP integration for both base and meta-models
  • 💾 Save/Load: Model, metrics, predictions, full stack
  • 🧰 CLI-ready with main.py and config.yaml
  • 📓 Notebook suite (01–07) for educational and production-ready use

📚 What's Inside

  • src/ecostackml/: modular codebase (data, models, preprocessing, utils)
  • notebooks/: seven tutorials, each focused on a pipeline stage
  • docs/: full documentation for MkDocs and GitHub Pages
  • main.py: launch the full pipeline from YAML
  • config.yaml: configure your entire workflow from one place
  • pyproject.toml: pip installable, PyPI-ready
  • LICENSE.md: MIT licensed, community friendly

🌐 Documentation

💬 A Word from the Author

EcoStackML was born out of the need for clarity and confidence in machine learning workflows — especially when applied to messy, real-world data.
Whether you're an environmental scientist, a student building your thesis, or an engineer shipping AI — this framework is for you.

Thank you to all collaborators, students, reviewers, and early testers.
The journey of EcoStackML has just begun 🚀

Assets 2
Loading