Tutorials and examples

Executable notebooks live in the repository examples/ directory on GitHub:

https://github.com/ecrl/graphchem/tree/master/examples

Property prediction

These notebooks load a bundled dataset, encode SMILES, and train or evaluate a MoleculeGCN:

• predict_cn.ipynb — cetane number

• predict_lhv.ipynb — lower heating value

• predict_mon.ipynb — motor octane number

• predict_ron.ipynb — research octane number

• predict_ysi.ipynb — yield sooting index

Comparison experiments

Additional training notebooks and CSV splits under examples/comparison/:

• train_cn.ipynb

• train_mon.ipynb

• train_ron.ipynb

Extras

Install optional example dependencies with:

pip install -e ".[examples]"