It would be nice to be able to consider the latitude and the longitude of the observations as a predictor.
Add latitude and longitude as computations in FER and FE. They are not computations, strictly speaking, there are not computations to do. Let's say the user selects the computation called "lat/lon", then no further actions or selection are required from the user (i.e. the user doesn't have to select any parameter or units or scaling factors) because the only thing that will happen is the latitude and longitude column (already available in the point data table) will be read as a predictor by the GUI. Those would be circular variables.
It would be nice to be able to consider the latitude and the longitude of the observations as a predictor.
Add latitude and longitude as computations in FER and FE. They are not computations, strictly speaking, there are not computations to do. Let's say the user selects the computation called "lat/lon", then no further actions or selection are required from the user (i.e. the user doesn't have to select any parameter or units or scaling factors) because the only thing that will happen is the latitude and longitude column (already available in the point data table) will be read as a predictor by the GUI. Those would be circular variables.