Data publication for Physics-informed Bayesian optimization of expensive-to-evaluate black-box functions

This publication provides the JCMsuite files and Python scripts required to reproduce the beam-splitter benchmark results from the paper titled "Physics informed Bayesian optimization of expensive-to-evaluate black-box functions" Requirements JCMsuite (tested with version 6.4.8) (https://jcmwave.com/) Python (tested with version 3.11) JCMoptimizer (tested with version 2.1.0) (https://optimizer.jcmwave.com/) In order to run the optimizations with JCMoptimizer and perform the finite-element simulations with JCMsuite you need to replace the corresponding placeholders with the paths to your local installations of JCMoptimizer and JCMsuite. Usage Running the script Benchmarking_optimizers_beam_splitter.py without modifying the directory structure will generate a plot equivalent to Fig. 7 of the manuscript. Note, however, that due to the inherent stochasticity of the optimization algorithms, the reproduced plot may differ slightly from the one shown in the paper. Notes We acknowledge fundings by Federal Ministry for Economic Affairs and Energy (BMWi, project number 50WM2253, AI-Quadrat), by the German Federal Ministry of Research, Technology and Space (BMFTR, project number 01IS24005, NanoMaC), and by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany´s Excellence Strategy – The Berlin Mathematics Research Center MATH+ (EXC-2046/1, project ID: 390685689).