Open josswright opened 4 years ago
Thanks for reporting and providing the appropriate patch. Behavior may have changed, but I don't know how this ran before. RStan can just read out of the environment, so that might've been happening somewhere before.
I just patched and added an acknowledgement.
I just went through the Lotka-Volterra example via the .Rmd, and also by pulling out the R code directly. There's a bug that seems to come from the creation of the
lynx_hare_data
list without named elements.With the current specification (line 527 of the
lotka-volterra-predator-prey.Rmd
):lynx_hare_data <- list(N, ts, y_init, y)
I get the following error when trying to fit the model:
This can be fixed by naming the list elements appropriately:
lynx_hare_data <- list(N=N, ts=ts, y_init=y_init, y=y)
I'm assuming that this must be a changed behaviour in some recent version of
rstan
?