search

rbchan / unmarked

R package for hierarchical models in ecological research
https://rbchan.github.io/unmarked/
36 stars 25 forks source link

Help for occuRN #109

Closed mikemeredith closed 6 years ago

mikemeredith commented 6 years ago

Suggested changes to the help file for occuRN to make it clearer. I know one postgrad who wasted a huge amount of time because he misunderstood it.

Main issue is notation, as the w matrix in Royle & Nichols (2003) this is the usual detection history, w[i,j] = 1 if the species is detected at siteiin samplej. In the help page w[i,j] indicates detection of an individual. In fact, since each individual (k) would have a different value, this would have to be w[i,j,k]. To make things worse, the actual detection matrix, for which you generally use y, doesn't appear in the help page.

I suggest taking out the reference to w, as we aren't interested in modelling detections of individuals, and include the model for the detection history.

mikemeredith commented 6 years ago

Sorry, slight mess there. The "Merge pull request #1..." thing is just updating my fork to match your master.

rbchan commented 6 years ago

Merged, thanks.

On Wed, Nov 8, 2017 at 10:44 PM, mikemeredith notifications@github.com wrote:

Sorry, slight mess there. The "Merge pull request #1 https://github.com/rbchan/unmarked/issues/1..." thing is just updating my fork to match your master.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/rbchan/unmarked/pull/109#issuecomment-343038965, or mute the thread https://github.com/notifications/unsubscribe-auth/AANoiuIxej0rPBU9jmrI7UWmB2aSxhrtks5s0nU3gaJpZM4QXV1y .

-- Richard Chandler Assistant Professor Warnell School of Forestry and Natural Resources University of Georgia 706-542-5815 chandlerlab.uga.edu