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
I am not sure whether it is just a memory problem, but when I run exp_smoothing() I never get any result. Tried the example from the help page and it still does not work even when reducing the data to n=20. Have to stop R from computing. Is this a known issue? I'm never getting any error messages. Using modeltime version 4.2.2 and R 4.0.2 on a laptop with Windows 10, 2.6 GHz CPU, 32 GB RAM.
The text was updated successfully, but these errors were encountered:
Hi Matt, thanks for the quick reply. I'm currently out of office and will send it next week. Good news is: I installed it on my personal notebook (v4.2.2 on R 4.0.2) with all packages from source and up to date and this time all works fine. So I guess either it's an bad installation or a function-conflict in some package.
Hi again, found out what caused the issue. Updating the forecast packages to v8.18 did the trick. The problem lied in function forecast::ets which you call in modeltime::ets_fit_impl. In v8.16 forecast::ets doesn't seem to find an appropriate solution while for v8.18 it converges within half a second. Here a snippet that worked in v8.18, but failed to converge in v8.16: timeseries_reprex.zip
I am not sure whether it is just a memory problem, but when I run exp_smoothing() I never get any result. Tried the example from the help page and it still does not work even when reducing the data to n=20. Have to stop R from computing. Is this a known issue? I'm never getting any error messages. Using modeltime version 4.2.2 and R 4.0.2 on a laptop with Windows 10, 2.6 GHz CPU, 32 GB RAM.
The text was updated successfully, but these errors were encountered: