Issue with prediction in svcPGOcc() when specifying grid.index #55
Description
v0.8.0 contains a bug in prediction for SVC occupancy models when using the grid.index functionality to specify SVCs at a larger spatial resolution from the actual point locations. When using predict(), a warning message would be displayed saying that there was a mismatch in the lengths of different vectors. This was a problem that resulted in inaccurate predictions when using svcPGOcc() and specifying the resolution of the SVCs as something higher than the point locations using the grid.index functionality. If you did this, I strongly encourage you to rerun the functions with v0.8.1, and I apologize for the inconvience this caused. Thanks very much to Ben Goldstein for pointing this out. Note that this bug did not affect results if using svcPGOcc() without specifying grid.index in the data list (which is likely the vast majority of instances for when svcPGOcc() is used).
This is now fixed in the development version here on GitHub (v0.8.1), but I will keep the issue open until v0.8.1 gets onto CRAN.