That's a great question. There may be a better way, but ultimately this is the fastest for the moment. Perhaps we can create an issue for this. It would be nice if we have that the sampling function considers, for instance, quarantine and other things that may affect the probability of sampling.
Originally posted by @gvegayon in #155 (comment)
See also #155 (comment)