Carlos Parada
Carlos Parada
Perhaps a long-term goal since it would require quite a bit of work and might involve some breaking changes, but I'd like it if bootstrap estimates had greater modularity and...
I'm not entirely sure what the balanced sampling bootstrap is doing -- there's several variance reduction techniques it could be referring to, and I'm not sure which one of these...
The Bayesian bootstrap simulates an approximation for the posterior distribution of a parameter by assuming that only the observed values are possible, and the relative frequency of each observed value...
I've built the [ParetoSmooth.jl](https://github.com/ParadaCarleton/ParetoSmooth.jl) package for fast approximate LOO cross-validation. The function currently only works with cross-entropy loss. (Bayesians usually call it expected log posterior density, and the package calls...
The current phrasing contains the same information but is a bit harder to interpret: ``` Computed from 4000 by 262 log-likelihood matrix ``` ParetoSmooth.jl is more explicit about this, with...
Only sort the tail values of the array, rather than the whole thing.
Implement plot from Bayesian workflow paper
[As described here](https://www.jstor.org/stable/2336308?seq=4#metadata_info_tab_contents); drawing from a Bayesian bootstrap, smoothed with a uniform interpolation between points, gives some [pretty big improvements](http://faculty.missouri.edu/kaplandm/pdfs/Kaplan_Hofmann_BB_quantile.pdf) to the quality of inferences.
I think a breakdown of LOO scores into calibration/refinement would be interesting and useful to users. Mentioning @avehtari because Jonah mentioned you'd proposed adding this decomposition and said you might...
The type signatures on everything make error messages very hard to read. Should we look at `@concrete terse struct` to cut down on that?