The Atomistic Cookbook

The cookbook contains recipes for atomic-scale modelling of materials and molecules, with a particular focus on machine learning and statistical sampling methods. Many of the examples rely heavily on software developed by the laboratory of computational science and modeling (COSMO, see its github page) but the cookbook is open for recipes using all types of modeling tools and techniques. Rather than focusing on the usage of a specific package, this cookbook provides examples of the solution of concrete modeling problems, often using a combination of several tools.

You can view the recipes online, compiled as webpages containing explanations, code snippets, plots and interactive viewers based on chemiscope. However, it is also possible (and hopefully simple) to download scripts, and conda environments, to run the recipe on your computer. You can use these as a starting point and a template that can be easily adapted to your own use case.

Table of contents

Recipe of the day

Want to try something new? Each day, one of the recipes in the cookbook is highlighted on the front page. There is one to suit everyone’s taste!

Hamiltonian Learning for Molecules with Indirect Targets

<p>This tutorial introduces a machine learning (ML) framework that predicts Hamiltonians for molecular systems. Another one of our cookbook examples demonstrates an ML model that predicts real-space Hamiltonians for periodic systems. While we use the same model here to predict a molecular Hamiltonians, we further finetune these models to optimise predictions of different quantum mechanical (QM) properties of interest, thereby treating the Hamiltonian predictions as an intermediate component of the ML framework. More details on this hybrid or indirect learning framework can be found in ACS Cent. Sci. 2024, 10, 637−648. and our preprint arXiv:2504.01187.</p>

This tutorial introduces a machine learning (ML) framework that predicts Hamiltonians for molecular systems. Another one of our cookbook examples demonstrates an ML model that predicts real-space Hamiltonians for periodic systems. While we use the same model here to predict a molecular Hamiltonians, we further finetune these models to optimise predictions of different quantum mechanical (QM) properties of interest, thereby treating the Hamiltonian predictions as an intermediate component of the ML framework. More details on this hybrid or indirect learning framework can be found in ACS Cent. Sci. 2024, 10, 637−648. and our preprint arXiv:2504.01187.