Graphics Reference
In-Depth Information
Figure
.
.
An estimate of the regression function of catch score on latitude and longitude, with colour
coding to add information on the relative size of the standard error of estimation across the surface
panel.hishighlightstheareas whereprecisionintheestimateislow.Itisnosurprise
tofind these at the edges of the surface, whereinformation is less plentiful. However,
it is particularly helpful to be able to identify the relatively high standard errors at
low latitude.
his idea extends to the assessment of reference models, such as a linear trend
acrosslatitude and longitude. hisisshown inthe right-hand panel of Fig.
.
,where
the surface panels are painted according to the size of the standardised difference
(
, to assess graphically the plausibility of a linear model p in two
covariates. he red panels indicate a difference of morethan
,and the blue panels of
less than
p
p
m
−
)
s.e.
(
m
−
)
, standard errors (s.e.).his gives an immediate graphical indication that
the curvature in the surface is not consistent with a linear shape.
Surfacesarethree-dimensionalobjectsandsotwaretodisplaysuchobjectsinhigh
quality is now widely available. he OpenGL system is a good example of this, and
access to these powerful facilities is now possible from statistical computing envi-
ronments such as
R
,throughthe
rgl
package described by Adler (
). he let
hand plot of Fig.
.
gives one example of this, showing a regression surface for the
Reef data with additional wire mesh surfaces to define a reference region for a lin-
ear model. he protrusion of the estimated surface through this reference region in-
dicates the substantial lack-of-fit of the linear model. his higher quality of three-
dimensional representation, together with the ability to rotate the angle of view in-
teractively, provides a very attractive and useful display.
−
