Appearance
Use Cases
SHOWME.how has been applied to three real computational studies, spanning biomedical engineering, geophysical mass flow simulation, and high-dimensional surrogate modelling.
Bayesian Calibration of a Hemolysis Model
Three collaborators, spread across macOS, Windows, and Linux, calibrate a hemolysis model against experimental measurements. A forward model in Python or Julia is queried by an MCMC unit in an online loop, with the study able to sweep across datasets or fan out across competing model implementations.
Bayesian Calibration of a Mass Flow Model
Two collaborators calibrate a geophysical mass flow simulator in two stages: a surrogate trained on simulator runs, then MCMC inference against that surrogate, across four units spanning Julia, Python, and R.
Benchmarking Gaussian Process Emulators for High-Dimensional Problems
A single researcher benchmarks twelve Gaussian process emulator implementations, in Python and R, against five datasets, fanning them out in parallel and aggregating the results, automated in both Nextflow and Snakemake.
Contributing a Use Case
Have you applied SHOWME.how to your own workflow? Contributions are welcome. A use case should include a description of the scientific context, the environment specs for each unit, a workflow definition, and enough sample data or a script to reproduce the results. Open a pull request against the study's own repository, then reach out so it can be added here.