posterior
posterior copied to clipboard
stan-dev
The posterior R package
The posterior package started with focus on multi-chain MCMC and stores chain and iterations ids. These are useful when computing multi-chain Rhat, ESS, and MCSE. It is also possible to...
For a very large number of variables stored in csv, it could be useful to use Arrow to read posterior csv as Arrow data table and use that to let...
Hi! rvars are a great thing and I have started using them a bit. I have run into issues when using it and there could be a nice solution but...
Not sure if this is a posterior issue or data.table issue but it looks like you cannot put rvars into data.table but it does work with data.frame ```R library(data.table) library(posterior)...
HDI?
Would it be in the scope of **posterior** to provide support for HDI's in contrast with quantile-based intervals? For example, say I wanted to quickly summarize my draws by their...
See https://github.com/stan-dev/posterior/issues/43 This is not ready for primetime, but I want the authors to be able to commit here if they want, so here it is. Now has: * A...
Pinging off of https://github.com/stan-dev/posterior/pull/157#issuecomment-846611937, would it be useful to have some standard benchmarks we use for evaluating performance? Does anyone have experience setting something like this up? I noticed that...
I was looking at what would be necessary to add parallelism to the computation of rstar, and thought this would work: ```r library(doMC) registerDoMC(4) rstar(x,uncertainty=T,trControl=caret::trainControl(allowParallel=T)) ``` But I see that...
When working on https://github.com/stan-dev/cmdstanr/issues/582 I was reminded that `variables()` and `nvariables()` behave differently for the `draws_rvars` format compared to all other formats. I don't think this difference (treating vector/matrix variables...
Copying @martinmodrak's [post from discourse](https://discourse.mc-stan.org/t/rvars-distinguishing-between-a-scalar-and-array-of-size-1/23726) here since it should probably be resolved in some way: > Hi, > it appears that the rvar interface in posterior cannot distinguish between a...