The notebooks contained here allows the reproduction of the results of the paper 'Probing atomic-scale symmetry breaking by rotationally invariant machine learning of multidimensional electron scattering, Mark P. Oxley, Maxim Ziatdinov, Ondrej Dyck, Andrew R. Lupini, Rama Vasudevan & Sergei V. Kalinin, npj Computational Materials volume 7, Article number: 65 (2021) ' https://www.nature.com/articles/s41524-021-00527-3 .
Please note that this notebook is highly modified from the original notebook to incorporate the latest version of atomai. This results in the slightly different convergence behavior to the published results so users will have to experiment to find the stopping points corresponding to the published results.
The experimental notebook should work for data output by the Swift software used on Nion UltraSTEMs. Users will have to modify the inputs and preprocessing to suit thier own data sets