when RMark and MARK results (seem to) differ | read first
Posted: Mon Feb 27, 2012 4:01 pm
Following is quoted from a recent posting by Jeff Laake, creator and maintainer of the RMark package. Please read and consider the following before posting anything that relates to 'differences in results between MARK and RMark':
If you get a difference between models constructed with RMark and MARK interface the reasons are:
1) data are different
2) design matrices are different
3) links are different
Even though it has been said many times, I'll repeat it. RMark is just a different interface and uses the same numerical software mark.exe to fit the models. If there are differences, check through the 3 reasons above. It will be one of them. A difference in links is quite common when running pre-defined models with MARK interface and comparing to the same in RMark. In particular parameters like pent and Psi have a default of an mlogit link in RMark but will get a sin link with the dot model in the pre-defined models in MARK interface. A dot model can be used with those parameters but the PIMS must be specified so they differ across the set of parameters that are summed to 1.
...
You can avoid the #1 reason above by using export.MARK within R and then use RMark import in the MARK interface. This will make sure that the data (including time intervals) and its structure are the same in the MARK interface. You should not get a difference between RMark run and re-running in MARK once you export into MARK interface. If you do that means there is a bug in the RMark interface like the one discovered recently.
-- jeff