Generalizing Ripley's K function to inhomogeneous populations

In spatial statistics, Ripley's K function (Ripley, 1977) is a classical tool to analyse spatial point patterns. Yet, it faces two major limits: it is only pertinent for homogeneous point processes and it does not allow the weighting of points.<br />We generalize it to get a new function, M, which oversteps these limits and detects spatial structures of inhomogeneous populations of weighted points