A FORTRAN program is presented for an algorithm to calculate Ripley's
K(r) estimator for n points within a study region with an irregularly
shaped boundary. When the boundary is digitized in vector mode, the re
sult is a polygon consisting of a sequence of m straightline segments
(sides). Each estimate of K(r) is computed from a single set of result
s obtained after n(n - 1)m operations. Examples of application are con
cerned with the clustering of wildcat wells in Canadian oil and gas pl
ays. It is shown that oil discovery wells in the Devonian Beaverhill L
ake Play, Central Alberta, are positively clustered. Strong deviations
from randomness are exhibited by wildcats and gas discoveries in the
Leduc Reef Complex-Windfall Play in the same region. The dataset for t
he latter example is relatively large, consisting of n = 249 points wi
thin a polygon with m = 264 sides.