Tutorials

There are a variety of scripts and notebooks to get you kicked off with using topsearch. All these can be found within the project github repository (https://github.com/IBM/topography-searcher/tree/main/examples), and we provide a short overview of these examples here.

Notebooks

There are three notebooks within the repository: example_function, dataset_roughness and atomic_cluster. We recommend starting with example_function as this notebook illustrates much of the key functionality of the package. In it we perform topographical analysis of a two-dimensional function, which allows direction visualisation of the algorithmic results.

The other two notebooks illustrate specific examples of topographical analysis to solve research challenges. We provide an example for exploring the structures and evolution of atomic clusters (atomic_cluster) and one for understanding the performance of machine learning models (dataset_roughness).

Scripts

Each of the notebooks listed above have equivalent Python scripts that can run in the command line. The notebooks are split across several scripts each, and there is an additional application of analysing the properties of molecules. The molecular folder contains examples for exploring properties of small molecules using quantum chemical methods. These examples are slow to run and have therefore only been included as scripts.