Some years ago I noticed that Strauss's seminal paper in Biometrika
(vol. 62, No. 2,
1975, pp. 467 - 475) states that there are n = 77 points in region I.
(See page 474,
just below Figure 1.) However if one looks at the region I pattern in
spatstat, e.g.:
X <- with(redwoodfull.extra, redwoodfull[regionI])
npoints(X)
one gets n = 71. A careful count of the points in Region I in Strauss's
Figure 1 (page 474)
also gives n = 71. A possible explanation for this anomaly is that
there are
hidden points in Strauss's Figure 1. These points would of necessity
lie very close to
points that we "know about" which fact would inflate the number of
r-close pairs, for
any value of r used in any modelling exercise.
It would be kind of nice to find out what is *really* going on, but I
expect that the
original data are sadly lost in the mists of time.
cheers,
Rolf Turner
On 12/05/13 09:01, Marcelino de la Cruz wrote:
You can complaint directly to to David Strauss for their unscientific
behaviour. :-p
As ?redwoodfull says:
"The dataset redwoodfull contains the full point pattern of 195 trees.
The window has been rescaled to the unit square. Its physical size is
approximately 130 feet across.
...
The plot of the data published by Strauss (1975) was scanned and
digitised by Sandra Pereira, University of Western Australia, 2002."
So you can digitize yourself the original true coordinates from the
plot in Strauss (1975).
Cheers,
Marcelino
El 04/12/2013 19:37, ldec...@comcast.net escribió:
does anyone know the original, true coordinates of the spatstat
redwoodfull data? i can't find them in:
Strauss, D. J. (1975). "A Model for Clustering." Biometrika 62(2):
467-475.
i think reporting spatial data in rescaled units is unscientific.
Lee De Cola, PhD, MCP.
DATA to Insight
ldec...@comcast.net
Reston, Virginia
703 709 6972
571 315 0577 mobile
_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-geo