Running the same set of models for multiple species

[benjamin]

Greetings,

I have a relatively large data set (>3,000 sites with multiple observations 3-22 per site) and have encounter histories for ~15 species at each of these sites. I know I can import an set of models through a previous analysis, and that seems to work fine. I am wondering if there is an efficient way (e.g., batch processing) of running the same set of models, with the same set of covariates, but for different encounter histories. Any suggestions would be welcomed. Thank you!

[darryl]

I'm pretty sure there has been some previous posts about batching (it can be done), although at some point you need to eyeball all the results to make sure there were no warnings. Don't just focus on the 'top' models either, as lower models may have had a problem that if corrected pushes them up the list.

[Bryan Wright]

If you know 'R' (or want to learn) then this sort of thing is easy using RMark (see Appendix C in "Gentle Intro").

-Bryan

[cooch]

If you know 'R' (or want to learn) then this sort of thing is easy using RMark (see Appendix C in "Gentle Intro").

-Bryan

True - except the question was posted about program PRESENCE (and in the PRESENCE part of the forum, although this isn't identified in the generated email), not MARK. RMark does not 'interface' with PRESENCE (although there have been *very* preliminary* discussions about pulling together something like an 'RPresence').

[dhewitt]

If you know 'R' (or want to learn) then this sort of thing is easy using RMark (see Appendix C in "Gentle Intro").

-Bryan

True - except the question was posted about program PRESENCE (and in the PRESENCE part of the forum, although this isn't identified in the generated email), not MARK. RMark does not 'interface' with PRESENCE (although there have been *very* preliminary* discussions about pulling together something like an 'RPresence').

True - but you can run almost all (all?) of the occupancy models that are in PRESENCE in MARK as well. And if you do it through RMark, the optimization is the only thing you need... so you could switch to RMark "fairly easily".

P.S. - Is there a way to get the forum emails to identify the forum section for the posting? That would be awesome.

[cooch]

If you know 'R' (or want to learn) then this sort of thing is easy using RMark (see Appendix C in "Gentle Intro").

-Bryan

True - except the question was posted about program PRESENCE (and in the PRESENCE part of the forum, although this isn't identified in the generated email), not MARK. RMark does not 'interface' with PRESENCE (although there have been *very* preliminary* discussions about pulling together something like an 'RPresence').

True - but you can run almost all (all?) of the occupancy models that are in PRESENCE in MARK as well.

Not quite all, but a sizeable fraction.

And if you do it through RMark, the optimization is the only thing you need... so you could switch to RMark "fairly easily".

The 'work' would be in translating structures that specify a model structure in PRESENCE into the equivalent in MARK.

P.S. - Is there a way to get the forum emails to identify the forum section for the posting? That would be awesome.

The new version of the forum (which I hope to roll out in January) will do that. Given the new forum is 'coming' (really), I'm not inclined to invest a huge amount in tweaking the current code base (although I admit it would be nice).

[egc]

P.S. - Is there a way to get the forum emails to identify the forum section for the posting? That would be awesome.

As per my earlier reply - this capability will be more or less 'built-in' in the new version of the forum. In the short-run, I've made a minor change to the current forum structure which should serve much the same purpose - albeit in a somewhat inelegant way. You'll be able to spot the 'change' with the next post (this one, in fact...).

[cooch]

P.S. - Is there a way to get the forum emails to identify the forum section for the posting? That would be awesome.

As per my earlier reply - this capability will be more or less 'built-in' in the new version of the forum. In the short-run, I've made a minor change to the current forum structure which should serve much the same purpose - albeit in a somewhat inelegant way. You'll be able to spot the 'change' with the next post (this one, in fact...).

Test to confirm its all working.