We begin by counting the number of RDP points that fall below a given maximum tolerance interval above $$d_{field}$$. When the number of these points reaches a fixed value, $$n_{fixed}$$, the point closest to $$d_{field}$$ is stored as the most probable radius for this iteration. This process is repeated increasing the tolerance around $$d_{field}$$ which results in another probable radius being stored, not necessarily the same found in the previous iteration. The final estimation for $$r_{cl}$$ and its associated error comes from averaging the set of radius values stored this way and taking its standard deviation, respectively.