Albitron-TPH2O: A Machine-Learning Plagioclase--Liquid Hygrothermobarometer for Joint T--P--H2O Estimation

Albitron-PTH2O 🌋🌋🌋 A Machine Learning-Based Geohygrothermobarometer with a User-Friendly GUI Welcome to the official repository for Albitron-PTH2O. This software provides a robust, machine learning-driven approach for petrologists and volcanologists to accurately estimate crystallization pressure (P), temperature (T), and melt water content (H2O) from mineral-melt equilibrium data. While traditional empirical models often struggle with complex, non-linear geological datasets, Albitron-PTH2O leverages advanced machine learning algorithms to improve predictive accuracy across various volcanic systems. ✨ Key Features High Accuracy: Trained on comprehensive experimental petrology datasets to robustly predict P-T-H2O conditions. User-Friendly GUI: No coding experience required. We provide standalone executable applications (for Windows) equipped with an intuitive graphical interface. Open Source & Reproducible: The core machine learning model and data processing scripts are fully accessible for transparent peer review and future community development. (Note: This software is currently under peer review at JP. A detailed methodology and case studies validating the model will be available in the upcoming publication.) 🔧Installation & Usage Using the Standalone GUI Application If you just want to use the tool without setting up a Python environment, you can download the pre-packaged application:1. Go to the **Releases** page of this repository (look at the right sidebar).2. Download the appropriate version for your operating system (`.exe` for Windows ). 3. Extract the downloaded file and double-click the application to launch the GUI.4. Load your dataset and start predicting