You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The variable names and and format of nuts-sampler parameters are incompatible with what plotting functions of the posterior package expect. (I hope I didn't overlook a function in posterior that deals with that..)
The methods sampler_diagnostics() returns an array where the names for chain, iteration and draw are ".chain" ".iteration" ".draw" and parameters in different arrays. However posterior expects a "long form" data.frame with variables names Iteration, Parameter, Value, Chain.
Its of course not a problem to reformat this myself, but I think it would be convenient if sampler_diagnostics() could also return data in a format the plotting functions of the posterior package can work with.
CmdStanR version number
cmdstanr_0.1.0
posterior version number
posterior_0.1.0
The text was updated successfully, but these errors were encountered:
Hey @gbiele, sorry for the delay! Just seeing this now.
I think you meant to refer to the bayesplot package instead of the posterior package, right? If so, then I think the latest bayesplot on github does what you need. There hasn't been a bayesplot CRAN release since I added it, but if you install bayesplot from github you can use the new methods for CmdStanMCMC objects that I added in stan-dev/bayesplot#227.
The variable names and and format of nuts-sampler parameters are incompatible with what plotting functions of the posterior package expect. (I hope I didn't overlook a function in posterior that deals with that..)
The methods
sampler_diagnostics()
returns an array where the names for chain, iteration and draw are ".chain" ".iteration" ".draw" and parameters in different arrays. However posterior expects a "long form" data.frame with variables names Iteration, Parameter, Value, Chain.Its of course not a problem to reformat this myself, but I think it would be convenient if
sampler_diagnostics()
could also return data in a format the plotting functions of the posterior package can work with.CmdStanR version number
cmdstanr_0.1.0
posterior version number
posterior_0.1.0
The text was updated successfully, but these errors were encountered: