Forum: developers
Monitor Forum | 
Start New Thread
| RE: xtab for abundance data [ Reply ] By: Peter Solymos on 2011-01-23 03:19 | [forum:3907] |
|
It is true that the mean would be more interpretable, but adding this option would trigger a chain of changes with unexpected bugs here and there. I can imagine adding an extractor option to xtab() function so it can return object$xtab / dim(object)[3] and maybe a presence-absence version. This is easy and virtually bug free. A cleaner thing would be to redesign the 3D (segmented) mefa classes, so those store the data as array (or list of sparse matrices with same dimensions, possibly keeping only one set of dimnames; also adding a data.frame to store segment specific details, so xtab would contain the array). This way, the current implementation (sum of segments) would be only one possible view of the data. This to come in mefa4? Needs careful planning ahead. |
|
| xtab for abundance data [ Reply ] By: Manuel Schneider on 2011-01-20 13:58 | [forum:3898] |
| As I understand xtab always contains the sum of all segments due to the use of xtabs(). This may not be always be practical for segments with abundance or presence/absence data. E.g. if segments are different observation times we may want the mean value in the xtab table. | |
