This new tutorial from Martin Modrák (above image by Modrák and Phil Clemson) is cool! It’s got slides, code, and exercises that you can do on your own to learn what simulation-based calibration is all about!
For tomorrow I have a post scheduled on some open problems in simulation-based calibration.
Thanks for the shout out! I’ll note that the SBC package underlying the tutorial (https://github.com/hyunjimoon/SBC) is currently mostly in the “minimum viable product” stage, but we plan to move slow and fix things :-) All the use cases we show are geared towards validating a model implementation, given that you trust your sampling algorithm, which is in contrast to how SBC was presented originally (validating an algorithm given that you trust your model implementation), but algorithm validation is still possible and we hope to get some case studies/vignettes on that as well.