points.geodata
Plots Spatial Locations and Data Values
This function produces a plot with points indicating the data locations. Arguments can control the points sizes, patterns and colors. These can be set to be proportional to data values, ranks or quantiles. Alternatively, points can be added to the current plot.
Usage
# S3 method for geodata
points(x, coords=x$coords, data=x$data, data.col = 1, borders,
pt.divide=c("data.proportional","rank.proportional",
"quintiles", "quartiles", "deciles", "equal"),
lambda = 1, trend = "cte", abs.residuals = FALSE,
weights.divide = "units.m", cex.min, cex.max, cex.var,
pch.seq, col.seq, add.to.plot = FALSE,
x.leg, y.leg = NULL, dig.leg = 2,
round.quantiles = FALSE, permute = FALSE, …)
Arguments
- x
a list containing elements
coords
anddata
described next. Typically an object of the class"geodata"
- a geoR data-set. If not provided the argumentscoords
anddata
must be provided instead.- coords
an \(n \times 2\) matrix containing coordinates of the \(n\) data locations in each row. Defaults to
geodata$coords
.- data
a vector or matrix with data values. If a matrix is provided each column is regarded as one variable or realization. Defaults to
geodata$data
.- data.col
the number of the data column. Only used if
data
is a matrix with columns corresponding to different variables or simulations.- borders
If an \(n \times 2\) matrix or data-frame with the coordinates of the borders of the regions is provided, the borders are added to the plot. By default it searches for a element named "borders" in the geodata object.
- pt.divide
defines the division of the points in categories. See
DETAILS
below for the available options. Defaults topt.divide = "data.proportional"
.- trend
specifies the mean part of the model. The options are:
"cte"
(constant mean - default option),"1st"
(a first order polynomial on the coordinates),"2nd"
(a second order polynomial on the coordinates), or a formula of the type~X
whereX
is a matrix with the covariates (external trend). If provided the trend is "removed" using the functionlm
and the residuals are plotted.- abs.residuals
logical. If
TRUE
and the value passed to the argumenttrend
is different from"cte"
the point sizes are proportional to absolute values of the residuals.- lambda
value of the Box-Cox transformation parameter. Two particular cases are \(\lambda = 1\) which corresponds to no transformation and \(\lambda = 0\) corresponding to the log-transformation.
- weights.divide
if a vector of weights with the same length as the data is provided each data is divided by the corresponding element in this vector. Defaults divides the data by the element
units.m
in the data object, if present, otherwise no action is taken and original data is used. The usage ofunits.m
is common for data objects to be analysed using the package geoRglm.- cex.min
minimum value for the graphical parameter
cex
. This value defines the size of the point corresponding the minimum of the data. Defaults to 0.5.- cex.max
maximum value for the graphical parameter
cex
. This value defines the size of the point corresponding the maximum of the data. Ifpt.divide = "equal"
it is used to set the value for the graphical parametercex
. Defaults to 1.5.- cex.var
a numeric vector with the values of a variable defining the size of the points. Particularly useful for displaying 2 variables at once.
- pch.seq
number(s) defining the graphical parameter
pch
.- col.seq
number(s) defining the colors in the graphical parameter
col
.- add.to.plot
logical. If
TRUE
the points are added to the current plot or image otherwise a display is open. Defaults toFALSE
.- x.leg, y.leg
x
andy
location of the legend as documented inlegend
.- dig.leg
the desired number of digits after the decimal point. Printing values in the legend uses
formatC
with argumentformat = "f"
.- round.quantiles
logical. Defines whether or not the values of the quantiles should be rounded. Defaults to
FALSE
.- permute
logical indication whether the data values should be randomly re-alocatted to the coordinates. See
DETAILS
below.- …
further arguments to be passed to the function
plot
, ifadd.to.plot = FALSE
; or to the functionpoints
, ifadd.to.plot = TRUE
.
Details
The points can be devided in categories and have different sizes
and/or colours according to the argument
pt.divide
. The options are:
- "data.proportional"
sizes proportional to the data values.
- "rank.proportional"
sizes proportional to the rank of the data.
- "quintiles"
five different sizes according to the quintiles of the data.
- "quartiles"
four different sizes according to the quartiles of the data.
- "deciles"
ten different sizes according to the deciles of the data.
- "equal"
all points with the same size.
- a scalar
defines a number of quantiles, the number provided defines the number of different points sizes and colors.
- a numerical vector with quantiles and length > 1
the values in the vector will be used by the function
cut
as break points to divide the data in classes.
For cases where points have different sizes the arguments
cex.min
and cex.max
set the minimum and the maximum
point sizes. Additionally,
pch.seq
can set different patterns for the points and
col.seq
can be used to define colors.
For example, different colors
can be used for quartiles, quintiles and deciles while a sequence of
gray tones (or a color sequence) can be used
for point sizes proportional to the data or their ranks.
For more details see the section EXAMPLES
.
The argument cex.var
allows for displaying 2 variables
at once. In this case one variable defines the backgroung colour
of the points and the other defines the points size.
The argument permute
if set to TRUE
randomly realocates the data in the coordinates.
This may be used to
contrast the spatial pattern of original data against another
situation where there is no spatial dependence (when setting
permute = TRUE
). If a trend
is provided the residuals
(and not the original data) are permuted.
Value
A plot is created or points are added to the current graphics device. A list with graphical parameters used to produce the plot is returned invisibily. According to the input options, the list has some or all of the following components:
the values of the quantiles used to divide the data.
the values of the graphics expansion parameter cex
.
the values of the graphics color parameter col
.
the values of the graphics pattern parameter pch
.
References
Further information on the package geoR can be found at: http://www.leg.ufpr.br/geoR.
See Also
plot.geodata
for another display of the data and
points
and plot
for information on the
generic R functions. The documentation of
par
provides details on graphical parameters.
For color schemes in R see gray
and
rainbow
.
Examples
# NOT RUN {
op <- par(no.readonly = TRUE)
par(mfrow=c(2,2), mar=c(3,3,1,1), mgp = c(2,1,0))
points(s100, xlab="Coord X", ylab="Coord Y")
points(s100, xlab="Coord X", ylab="Coord Y", pt.divide="rank.prop")
points(s100, xlab="Coord X", ylab="Coord Y", cex.max=1.7,
col=gray(seq(1, 0.1, l=100)), pt.divide="equal")
points(s100, pt.divide="quintile", xlab="Coord X", ylab="Coord Y")
par(op)
points(ca20, pt.div='quartile', x.leg=4900, y.leg=5850)
par(mfrow=c(1,2), mar=c(3,3,1,1), mgp = c(2,1,0))
points(s100, main="Original data")
points(s100, permute=TRUE, main="Permuting locations")
## Now an example using 2 variable, 1 defining the
## gray scale and the other the points size
points.geodata(coords=camg[,1:2], data=camg[,3], col="gray",
cex.var=camg[,5])
points.geodata(coords=camg[,1:2], data=camg[,3], col="gray",
cex.var=camg[,5], pt.div="quint")
# }