Closed jgabry closed 4 years ago
I have some very basic implementations called read_optim_csv
and read_vb_csv
in R/utils.R
.
Just need to replace rstan::read_stan_csv
now
How does this differ from #2 ? I might just be missing something.
We can close this now though since I think we just need to read the csv for sampling now and that is covered by #2
The csv output is different depending on whether the CmdStan
method
is sample, optimize or variational. Right nowrstan::read_stan_csv()
is used formethod=sample
, but it doesn't work for the csv output for optimization or variational inference. We also don't want to depend on RStan (#2), so we should eventually not rely on it formethod=sample
either.