Florian Hartig

ok, this was probably accidentally committed to this branch after the merge (branch should have been deleted). It's not really crucial, and we could change it by hand, but just...

Hi Marco, yeah, we should fix the automatic thing, but you can do everything you want by hand. Note that in nearly all BT functions, the ... argument is forwarded...

sure, that is how I would have understood the request, but we don't have anything that I would finde sensible to cite. I'll put this on won't fix for the...

Is it necessary to thin after the sampler has run? Why don't you thin the sampler directly (option thin?)

Yes, but if you thin them from the start, you shouldn't have the memory problems you describe. I understand that an adaptive thinning might be a bit more convenient, but...

Hi Marjolein, 100 parameters is usually not large enough to cause memory problems. How many iterations are you doing here, and what's the size of your MCMC sampler object? Can...

Hi @marjoleinbruijning, OK, I see, this is the culprit. I have been meaning to look at the z matrix for a while, because I think in general it would be...

I don't really know any more what the problem was, this works with the old BT version ``` ## Generate a test likelihood function. ll

Ideally would be to create a general script for benchmarking, i.e. a number of fixed test functions etc., that we can use to benchmark different samplers. I will send you...

Yes, this should be possible. See, e.g., option "blockUpdate" in DEzs.