Multiscale occ: categorical variables for theta
Posted: Thu Jan 05, 2017 1:46 pmHi folks,
I'm using the multiscale occupancy model in RMark: MultScalOcc. I have a bunch of sites and each site contains many plots, so I'm estimating psi for sites and theta for plots.
I'm coding theta (plot) level variables as time-varying covariates, like cover1, cover2, cover3...where the number corresponds to the plot. This works well for continuous variables but I also have categorical plot variables.
Is it possible to use categorical variables to model theta?
Is it possible to format a categorical variable as a time-varying covariate?
My attempts so far have failed. I've got the column names for the categorical variable setup to be time-varying but the columns are factors, so that causes multiple problems, of course. I could setup the 1/0 dummy coding for the variable but then there would be multiple columns for each plot and I don't understand how that would work. The variable has 5 levels.
Thanks!
Joe
I'm using the multiscale occupancy model in RMark: MultScalOcc. I have a bunch of sites and each site contains many plots, so I'm estimating psi for sites and theta for plots.
I'm coding theta (plot) level variables as time-varying covariates, like cover1, cover2, cover3...where the number corresponds to the plot. This works well for continuous variables but I also have categorical plot variables.
Is it possible to use categorical variables to model theta?
Is it possible to format a categorical variable as a time-varying covariate?
My attempts so far have failed. I've got the column names for the categorical variable setup to be time-varying but the columns are factors, so that causes multiple problems, of course. I could setup the 1/0 dummy coding for the variable but then there would be multiple columns for each plot and I don't understand how that would work. The variable has 5 levels.
Thanks!
Joe