Hi, I have a circular shaped set of point on the plane (X,Y) centered in zero. The distribution is more dense close to zero and less dense far from zero. I need to find the radius of a circle centered in zero that contains 65% of the points in the sample. Is there any R directive that can do this? I wanna start with 2D set of points, but the real case scenario is with a 5D set of points. Thanks, Francesco
confidence interval
2 messages · Francesco Petrogalli, David Winsemius
On Jan 21, 2011, at 11:33 AM, Francesco Petrogalli wrote:
Hi, I have a circular shaped set of point on the plane (X,Y) centered in zero. The distribution is more dense close to zero and less dense far from zero. I need to find the radius of a circle centered in zero that contains 65% of the points in the sample. Is there any R directive that can do this? I wanna start with 2D set of points, but the real case scenario is with a 5D set of points.
Something along the lines of dxy= with(dfm, sqrt(x^2 +y^2)) quantile(dxy, probs=0.65) The generalization to 5 dimensions appears trivial. Even the generalization to finding the radius around an arbitrary point seems trivial assuming an L2 norm.
David Winsemius, MD West Hartford, CT