Matthew Stephens

potential to add check that data files are committed

4

How feasible would it be to add a check that any files being sourced or loaded are included in the repo? eg. it seems to be an easy mistake to...

when i rename an rmd file using `wflow_rename("example.rmd")` the corresponding html file gets renamed (great) and so do the figures (great) but the html in the html file does not...

docs on how bpipe handles dependencies

3

I've read through several parts of the documentation and tutorial, but couldn't find a description of how dependencies are handled. If an input file is changed, does bpipe rerun that...

i think we could do with some more vignettes eg for EE vs ET and using moderated t: https://github.com/stephenslab/ash/blob/master/analysis/CompareEEvsETmodel.Rmd https://github.com/stephenslab/EBNM-moderated-t/blob/master/analysis/method.Rmd

diagnostic plot

8

would be good to produce a diagnostic plot. For each element betahat_j compute where it falls (ie its quantile) in its predictive distribution p( betahat_j | ghat, s_j) If the...

bug in my_etruncnorm

5

there seems to be a bug here: > ashr:::my_etruncnorm(0,99,-2,3) [1] 5.795534 > truncnorm:::etruncnorm(0,99,-2,3) [1] 1.795534

ashCI doesn't work when df varies across observations

2

this might be useful to have...

would be nice to implement something to compute posterior medians

default prior for general likelihoods

1

@mengyin for general likelihoods does it make sense to have the "null-biased" prior as default? Might be better to make it uniform? What is your experience?

the current computation of output is nice and modular, but also much slower than it needs to be. Possible speeds ups would involve rearranging so that i) we only compute...