Yep, just about what one would expect from randomly cutting the top of your uniform distribution off multiple times! What's interesting is the transforms used

Yep, just about what one would expect from randomly cutting the top of your uniform distribution off multiple times! What's interesting is the transforms used to sample in some new space from e.g. uniform random inputs. For instance, disks [1] or hemispheres [2] or an arbitrary density function [3]. Common stuff in light transport simulation, and easy to get wrong. [1]: https://stats.stackexchange.com/questions/481543/generating-random-points-uniformly-on-a-disk https://stats.stackexchange.com/questions/481543/generating-... [2]: https://alexanderameye.github.io/notes/sampling-the-hemisphere/ https://alexanderameye.github.io/notes/sampling-the-hemisphe... [3]: https://en.wikipedia.org/wiki/Inverse_transform_sampling https://en.wikipedia.org/wiki/Inverse_transform_sampling