AR1 function sd estimate wrongly scaled

[TahminaMojumder]

likelihoodAR1 parameter sd actually fits (1-a) * sd

[florianhartig]

So, this reference https://faculty.washington.edu/ezivot/econ584/notes/armaestimation.pdf gives the following formula for the likelihood

Check

• If this is identical to the code
• Can we have a citable reference for this, e.g. from a reputable book?

[TahminaMojumder]

https://www.google.de/books/edition/Applied_Time_Series_Analysis/9sM3Lpxgr7EC?hl=en&gbpv=1&dq=unconditional+log+likelihood+for+ar1&pg=PA250&printsec=frontcover

[florianhartig]

OK, I think this wasn't resolved yet but we agreed the scaling is wrong and we have to change either help or (preferred) the code?