Examples

Run these examples from the repository root after installing the package in the required environment. The examples use tiny local data, keep optional services disabled by default, and only the TOML training workflow writes persistent outputs under example_outputs/.

• Feature Engineering
• Tree To Graph
• Tree I/O
• Single-Task Training
• TOML Training Configuration
• Complete Pipeline

Recommended order

1. Feature Engineering Attaches node features to a small in-memory tree and prints a node-by-node feature summary.

2. Tree To Graph Converts a featured tree to graph tensors, then demonstrates virtual time-bin nodes.

3. Tree I/O Loads a repository sample tree when optional DendroPy dependencies are installed, or exits cleanly with installation guidance.

4. Single-Task Training Runs a compact single-task training loop with temporary output cleanup.

5. TOML Training Configuration Trains a small configured model and writes a checkpoint and history under example_outputs/toml_training_config/.

6. Complete Pipeline Builds a tree, attaches features, converts it to graph data, loads a checkpoint or creates a temporary one, and prints a prediction.

Each page is a script launcher with inputs, run command, stable stdout markers, files written, optional dependencies, failure modes, and source.