about waicAbund

[QAQ0w0]

I have noticed that WAIC calculations are slow, sometimes even exceeding the model fitting time. Is there any way to improve the calculation speed

[doserjef]

Hello,

Thanks for the note! As you point out, waicAbund() can be slow, particularly for N-mix and distance sampling models. It will also be slower the larger your values of abundance are. It should be much faster when working with GLMMs. The slowness is due to the need to integrate out the latent abundance values in N-mix and distance sampling models, which I decided to implement separately from the actual model fitting functions as opposed to doing that integration directly in the model-fitting functions (which would then make the call to waicAbund() much faster but the actual model-fitting function slower). There's nothing that can really be done on the user side to speed the function up at this point. Sorry about that. At some point I would like to try to dig into ways to make it a bit faster (most likely through parallelization), but that probably won't be for a while. I'll go ahead and leave this open for now.

Jeff