Use of Bayesian Inference to Diagnose Issues in Experimental Measurements of Mechanical Disk Resonators - Inference Results

Data release for Use of Bayesian Inference to Diagnose Issues in Experimental Measurements of Mechanical Disk Resonators. This record contains results for all the inference analyses perfomed in the paper. Visualizing results The accompanying data release page includes instructions for how to read the results provided in the data release and produce the figures include in the paper. Directory structure The ZIP file includes results for analyses using real and simulated data and has the following file structure: - bayesbeat_inference_results
- real_data
- model_1
- model_1_constant_noise
- model_3
- simulated_data
- model_1
- model_1_constant_noise
- model_3 Each directory is labelled with the model that was used to perform that analyses and contains result files for each format that using the following naming convention: result_ringdown_.hdf5 where ID denotes which ringdown the result corresponds to. File format The included files are result files produced by the nested sampler nessai. These are saved using the HDF5 file format and can read using the h5py Python package. Each file includes various fields, including: Posterior samples A log evidence estimate The total run time and number of likelihood evaluations