Add two example scripts to demonstrate how Bayesian optimization with Gaussian processes finds the experimental parameters that minimize data error after reconstructions.
Both examples utilize a wrapper function (ptycho_recon_exp_data_electron) that carries out ptychographic reconstructions based on user-speficied experimental parameters.
Parallel BO is highly recommended, especially for multislice reconstructions due to their long computational time.
As an example, here is a BO run for tuning probe defocus and sample thickness of the PSO data.
Add two example scripts to demonstrate how Bayesian optimization with Gaussian processes finds the experimental parameters that minimize data error after reconstructions.