Lukas Lohse

Hi, I also took a look at this and while trying to more closely match the original analysis I found the issue seemingly stemming from theta being estimated as too...

Fascinating, the degrees of freedom in the performance package don't account for the number of groups. No matter if you are using 2, 10, or 1000 groups the df are...

Hi michielvdglind, I've been spending some time double checking the p-value calculations in this package, so i feel i'm qualified to somewhat answer your question. Let's look at what bolker...

Hi Florian, sorry for the confusion earlier. I did some simulations to check my intuitions. The Code could be turned into functions, but whatever. (sorry for the .txt-file, Github discriminates...

Hi Florian, i guess 3 weeks counts as soonish? Well, everyone's lives are difficult right now and i've wasted a substantial amount of time planning a vacation that never happened....

Sorry for the delay. I spent some more time thinking and i think i figured out, why my test idea is slightly anti-conservative. Since the PIT is performed with an...

Hi, no worries, take your time. I know i did and i won't have much time this weekend anyway. Just a short response to your initial thoughts. Issue 2 and...

I have a few issues with this. Admittedly i haven't been keeping up with the development version of DHARMa to not disrupt my old code. So our results might differ....

Hi Florian, I've thought a little about REs overfitting and i don't think it's "real". Random coefficients are still a type of residual, just transformed to the scale of the...

Hello Florian, i'd like to do something productive on the upcoming weekend and i figured this would be a good place to start. However before i do, it would be...