The SPP Procedure

Ripley-Rasson Window Estimator

When the sampling window for a point pattern is unknown, it can be estimated from the data. A common error is to assume that $W$ is the smallest rectangle that contains the data points, or is the convex hull that encloses the data points (Baddeley, 2010). Either choice is an underestimate of the true region $W$ and usually yields an overestimate of the point process intensity $\lambda $ or summary statistics such as the K function. The Ripley-Rasson window estimator is an estimate of the spatial window from which the points were drawn (Ripley and Rasson, 1977). For estimating a rectangular study region, the Ripley-Rasson estimate is the rescaled copy of the minimum bounding box of the collection of points, centered at the box’s centroid and expanded using a scaling factor of $1 / \sqrt {1-\frac{4}{n}}$, where n is the number of data points.