A properly implemented ORS should in theory be much faster than the standard rejection sampler. This wasn't the case with the previous implementation of ORS. That ORS is faster than RS has already been observed in the python version of the code, and most probably will make the injection process much more efficient.
A properly implemented ORS should in theory be much faster than the standard rejection sampler. This wasn't the case with the previous implementation of ORS. That ORS is faster than RS has already been observed in the python version of the code, and most probably will make the injection process much more efficient.