The function SamplingBias of sampbias package (https://github.com/azizka/sampbias) have an argument to specify the sample distributions, the "binsize".

"Indicating the distance bin size (in meters!) for approximating the empirical and observed sample distributions. Should be slightly larger than res (which is in degrees). Default = res * 100000 * 1.1."

I have doubts how to refine this argument to my question.

I would like to evaluate the sampling biases of carnivores parasitological samples points pattern in Brazil. I have used the default list of geographic gazetteers and others SpatialPolygonsDataFrame.