backgroundRadius fails for large datasets

[lukasbaumbach]

The backgroundRadius runs fine for your small example data set. However, for larger spatial extents (say 1000 rows x1000 columns) these calculations quickly become very memory intense due to the inefficient computation behaviour of matrices. In my example I got prompted, that the matrix is too large and even when installing the recommended spam64 package, I get the error that the matrix distances are too dense.

Since most of the previous steps work with rasters, is there maybe a way to translate your function to work on the basis of rasters? This may save a lot of workspace (in my example R crashed, since the working memory of 32GB was exceeded) and computation time!