mltsa Documentation

mltsa is the new public package for Machine Learning Transition State Analysis. The current documentation focuses on the modern src/mltsa API, the reusable HDF5 workflows, and the migration path from the historical repository layout.

Guide

• Installation
  • Base install
  • MD extras
  • Notes
• Package Overview
  • Package areas
  • Repository state
  • Storage model
• Synthetic Datasets
  • Main entry points
  • Typical workflow
  • What is stored
• Models and Explainability
  • Model factory
  • Supported wrappers
  • Explainability
  • Available methods
• MD Workflow
  • 1. Label trajectories
  • 2. Build feature sets
  • 3. Train and explain
  • 4. Export structures
  • Minimal example
• CLI Usage
  • Root command
  • MD commands
  • Label trajectories
  • Build a feature set
  • Run analysis
  • Keyword arguments
• Notebooks
  • Current tutorials
  • Legacy material
  • Migration note
• Upgrade Guide
  • Concept map
  • Typical migration patterns
  • What stays legacy for now