rayplot(x, y, z, lowbnd, highbnd, minz = <<see below>>, maxz = <<see below>>, arcs = list(length=0.75, sides=18, lwd=1, density=-1, border=F, col=2), tics = list(length=0.06, sides= 8, lwd=1, col=1), dots = list(length=0.06, sides= 8, lwd=1, density=0, border=T, col=1), rays = list(length=0.30, lwd=2, col=1), minangle= -pi/2, maxangle=pi/2, clockwise=F)
z.
If provided then it must be the same length as
z.
z.
Must be the same length as
z, if provided. Either both or none of
lowbnd and
highbnd must be given.
z by.
minz maps into
minangle so
only one of the two is necessary. Its default value is
min(z) if
lowbnd is missing or
min(lowbnd) otherwise.
z by,
maxz maps into
maxangle so
only one of the two is necessary. Its default value is
max(z) if
highbnd is missing or
max(highbnd) otherwise.
lowbnd and
highbnd are provided. The possible components of this list
are:
length= numeric scalar or vector giving the confidence sector radii in inches.
If a constant, then it is taken to be a fraction of the ray length.
If a vector (of the same length as
x), then each component is taken to
represent the corresponding arc's length.
sides= scalar giving the number of polygon sides for the maximum sector.
lwd= sector line width
density= sector fill options. If
density is zero, no filling will occur.
If density is negative, the sector will be filled solidly using the
device-dependent polygon filling algorithm.
border= logical flag. Should the border of the sector be plotted?
col= integer determining the color for the sector.
length=
radius of the tics in inches. No tics are plotted if this is
0 or
FALSE.
sides=
number of tics at the base of a ray. Default is
8.
lwd= tic line width.
col=
integer determining the color of the tics.
length=
numeric scalar or vector giving the radius of the dot in inches.
No dots are plotted if this is
0 or
FALSE.
sides=
integer. Number of sides of the "dot."
lwd=
line width parameter.
density=
dot fill parameter. See its definition above for
arcs.
border=
logical flag. Should the border of the dot be plotted?
col=
integer determining the color for the dot.
rays affect the characteristics of the main rays:
length=
numeric scalar or vector giving the ray length in inches. No rays are plotted
if this is
0 or
FALSE. If this is a scalar, it will be replicated to the
a length equivalent to the number of points at which rays are desired,
length(x).
lwd=
line width parameter.
col=
integer determining the color of each ray.
-pi/2 and the greatest to
pi/2.
Only one direction can be mapped with the current release, if both directions
are desired, then call
rayplot twice setting
clockwise
to
FALSE and
TRUE respectively.
Confidence sectors are useful but should be used with caution. Big arcs
will call attention to the less accurate estimates. Consequently, the user may
want to use shorter arc lengths for the less accurate estimates to de-
emphasize this effect.
The options
rays affects the characteristics of the main rays. If the
given
length for these is a constant value, it will be replicated to the
a length equivalent to the number of points at which rays are desired.
The option for
tics provides reference angles against which the user can
read the ray angles more accurately.
The option for
dots will put a dot at the base of each ray. Typically,
an open dot surrounds the tics.
Outliers are not easily spotted when encoded as ray-angles. If the user
wants to highlight selected observations or determine global scaling,
rayplot
should be called for each of the selected data subsets with changing
parameters, such as color (
col), length (
length), or line width (
lwd).
Carr, D. B., Olsen, A. R. and White, D. (1992). Hexagon mosaic maps for display of univariate and bivariate geographical data. Cartography and Geographics Information Systems 19, 228-236.
# Random locations values and bounds
x <- rnorm(100)
y <- rnorm(100)
z <- rnorm(100)
inc <- abs(rnorm(100,sd=.4))
lowbnd <- z-inc
highbnd <- z+inc
# No confidence bounds
plot(x,y,type='n')
rayplot(x, y, z)
# Confidence bounds
plot(x,y,type='n')
rayplot(x, y, z, lowbnd, highbnd)
# Clockwise orientation
plot(x,y,type='n')
rayplot(x, y, z, lowbnd, highbnd, clockwise=TRUE)
# No tics and small filled dots
plot(x,y,type='n')
rayplot(x, y, z, lowbnd, highbnd, tics=list(F),
dots=list(sides=20, length=.025, density=-1))
# Bivariate rays for smoothed data on a hexagon grid
# (Data not provided)
plot(mymap,type='l')
rayplot(grid$x,grid$y,pred$so4)
rayplot(grid$x,grid$y,pred$no3,clockwise=T,tics=list(F))