[R-sig-Geo] Warnings: " Window containing simulated patterns is not a subset of data window"

Rolf Turner r.turner at auckland.ac.nz
Fri Jan 21 04:27:27 CET 2011


I had a look just now at the data that you sent me.

The problem is pretty obvious --- the window for the ``data''
pattern is not the same as the window for the ``landscape''
pattern.  It is a proper sub-window.

Since the pattern ``data'' is a sub-pattern of the pattern
``landscape'' it is mysterious as to why you changed the
window for ``data'' to something smaller.  Be that as it
may, if you do

	data$window <- landscape$window

then your envelope calculations run without producing any
warnings.

The simulated patterns produced within envelope() will be
patterns on the window corresponding to ``intens'' which
is in turn the window of ``landscape''.  It doesn't really
make sense to compare the (inhomogeneous) K function of
a pattern living in window W_1 with those of patterns living
in window W_2 unless W_1 = W_2.

	cheers,

		Rolf Turner



More information about the R-sig-Geo mailing list