mapMM: Colorful plots of predicted responses in two-dimensional space.

Description

These functions provide either a map of predicted response in analyzed locations, or a predicted surface. mapMM is a straightforward representation of the analysis of the data, while filled.mapMM copes with the fact that all predictor variables may not be known in all locations on a fine spatial grid, but may involve questionable choices as a result (see map.formula argument). Both functions takes an HLfit object as input. mapMM calls spaMMplot2D, which is similar but takes a more conventional (x,y,z) input.

Usage

spaMMplot2D(x, y, z,xrange=range(x, finite = TRUE),
      yrange=range(y, finite = TRUE),
      margin=1/20,add.map= FALSE, nlevels = 20, 
      color.palette = spaMM.colors,map.asp=NULL,
      col = color.palette(length(levels) - 1), 
      plot.title=NULL, plot.axes=NULL, decorations=NULL,
      key.title=NULL, key.axes=NULL, xaxs = "i", 
      yaxs = "i", las = 1, axes = TRUE, frame.plot = axes, ...) 
mapMM(fitobject,Ztransf=NULL,coordinates,
      add.points,decorations=NULL,plot.title=NULL,plot.axes=NULL,envir=-3, ...)
filled.mapMM(fitobject, Ztransf = NULL, coordinates, xrange = NULL,
             yrange = NULL, margin = 1/20, map.formula, phi =
             1e-05, gridSteps = 41, decorations =
             quote(points(pred[, coordinates], cex = 1, lwd = 2)),
             add.map = FALSE, axes = TRUE, plot.title = NULL,
             plot.axes = NULL, map.asp = NULL, variance = NULL,
             var.contour.args = list(), smoothObject = NULL, ...)

Arguments

fitobject

The return object of a corrHLfit call.

x,y,z

Three vectors of coordinates, with z being expectedly the response.

Ztransf

A transformation of the predicted response, given as a function whose only required argument can be a one-column matrix. The name of this argument must be Z (not x), as is appropriate for use in do.call(Ztransf,list(Z=Zvalues)).

coordinates

The geographical coordinates. By default they are deduced from the model formula. For example if this formula is resp ~ 1 + Matern(1| x + y ) the default coordinates are c("x","y"). If this formula is resp ~ 1 + Matern(1| x + y + z ), the user must choose two of the three coordinates.

xrange

The x range of the plot (a vector of length 2); by default defined to cover all analyzed points.

yrange

The y range of the plot (a vector of length 2); by default defined to cover all analyzed points.

margin

This controls how far (in relative terms) the plot extends beyond the x and y ranges of the analyzed points, and is overriden by explicit xrange and yrange arguments.

map.formula

Plotting a filled contour generally requires prediction in non-oberved locations, where predictor variables used in the original data analysis may be missing. In that case, the original model formula cannot be used and an alternative map.formula must be used to interpolate (not smooth) the predicted values in observed locations (these predictions still resulting from the original analysis based on predictor variables). As a result (1) filled.mapMM will be slower than a mere plotting function, since it involves the analysis of spatial data; (2) the results may have little useful meaning if the effects of the original predictor variables is not correctly represented by this interpolation step. For example, it may involve biases analogous to predicting temperature in non-oberved locations while ignoring effect of variation in altitude in such locations.

phi

This controls the phi value assumed in the interpolation step. Ideally phi would be zero, but problems with numerically singular matrices may arise when phi is too small.

gridSteps

The number of levels of the grid of x and y values

variance

Either NULL, or the name of a component of prediction variance to be plotted. Must name one of the components that can be returned by predict.HLfit. variance="predVar" is suitable for uncertainty in point prediction.

var.contour.args

A list of control parameters for rendering of prediction variances. See contour for possible arguments (except x, y, z and add).

add.map

Either a boolean or an explicit expression, enclosed in quote (see Examples). If TRUE, the map function from the maps package (which much therefore the loaded) is used to add a map from its default world database. xrange and yrange are used to select the area, so it is most convenient if the coordinates are longitude and latitude (in this order and in standard units). An explicit expression can also be used for further control.

levels

a set of levels which are used to partition the range of z. Must be strictly increasing (and finite). Areas with z values between consecutive levels are painted with the same color.

nlevels

if levels is not specified, the range of z, values is divided into *approximately* this many levels (a call to pretty determines the actual number of levels).

color.palette

a color palette function to be used to assign colors in the plot.

map.asp

the y/x aspect ratio of the 2D plot area (not of the full figure including the scale). By default, the scales for x and y are identical unless the x and y ranges are too different. Namely, the scales are identical if (plotted y range)/(plotted x range) is 1/4 < . < 4, and map.asp is 1 otherwise.

col

an explicit set of colors to be used in the plot. This argument overrides any palette function specification. There should be one less color than levels

plot.title

statements which add titles to the main plot. See Details for differences between functions.

plot.axes

statements which draw axes (and a box) on the main plot. See Details for differences between functions.

decorations

Either NULL or Additional graphic statements (points, polygon, etc.), enclosed in quote (the default value illustrates the latter syntax). .

add.points

Obsolete, use decorations instead.

envir

Controls the environment in which plot.title, plot.axes, and decorations are evaluated. mapMM calls spaMM2Dplot from where these graphic arguments are evaluated, and the default value -3 means that they are evaluated within the environment from where mapMM was called.

key.title

statements which add titles for the plot key.

key.axes

statements which draw axes on the plot key.

xaxs

the x axis style. The default is to use internal labeling.

yaxs

the y axis style. The default is to use internal labeling.

las

the style of labeling to be used. The default is to use horizontal labeling.

axes, frame.plot

logicals indicating if axes and a box should be drawn, as in plot.default.

smoothObject

Either NULL, or an object inheriting from class HLfit (hence, an object on which predict.HLfit can be called), predicting the response surface in any coordinates. See Details for typical usages.

…

further arguments passed to or from other methods. For mapMM, all such arguments are passed to spaMMplot2D; for spaMMplot2D, currently only additional graphical parameters passed to title() (see Details). For filled.mapMM, these parameters are those that can be passed to spaMM.filled.contour.

Value

filled.mapMM returns invisibly a predictor of the response surface. mapMM has no return value. Plots are produced as side-effects.

Details

The smoothObject argument may be used to redraw a figure faster by recycling the predictor of the response surface returned invisibly by a previous call to filled.mapMM.

For smoothObject=NULL (the default), filled.mapMM interpolates the predicted response, with sometimes unpleasant effects. For example, if one interpolates probabilities, the result may not be within [0,1], and then (say) a logarithmic Ztransf may generate NaN values that would otherwise not occur. The smoothObject argument may be used to overcome the default behaviour, by providing an alternative predictor.

If you have values for all predictor variables in all locations of a fine spatial grid, filled.mapMM may not be a good choice, since it will ignore that information (see map.formula argument). Rather, one should use predict(<fitobject>,newdata= <all predictor variables >) to generate all predictions, and then either spaMM.filled.contour or some other raster functions.

The different functions are (currently) inconsistent among themselves in the way they handle the plot.title and plot.axes argument:

spaMM.filled.contour behaves like graphics::filled.contour, which (1) handles arguments which are calls such as title(.) or {axis(1);axis(2)}; (2) ignores ... arguments if plot.title is missing; and (3) draws axes by default when plot.axes is missing, given axes = TRUE.

By contrast, filled.mapMM handles arguments which are language expressions such as produced by quote(.) or substitute(.) (see Examples).

mapMM can handles language expressions, but also accepts at least some calls.

Examples

Run this code

# NOT RUN {
data("blackcap")
bfit <- corrHLfit(migStatus ~ means+ Matern(1|longitude+latitude),data=blackcap,
                  HLmethod="ML",
                  ranFix=list(lambda=0.5537,phi=1.376e-05,rho=0.0544740,nu=0.6286311))
if (requireNamespace("maps")) { ## required for add.map=TRUE 
  mapMM(bfit,color.palette = function(n){spaMM.colors(n,redshift=1/2)},add.map=TRUE)
}

if (spaMM.getOption("example_maxtime")>0.8) {
  ## filled.mapMM takes a bit longer
  # showing 'add.map', 'nlevels', and contour lines for 'variances'
  if (requireNamespace("maps")) { ## required for add.map=TRUE 
    filled.mapMM(bfit,nlevels=30,add.map=TRUE,plot.axes=quote({axis(1);axis(2)}),
             variance="respVar",
             plot.title=title(main="Inferred migration propensity of blackcaps",
                               xlab="longitude",ylab="latitude"))
  }
}

if (spaMM.getOption("example_maxtime")>2.5) {
 data("Loaloa")  
 lfit <- corrHLfit(cbind(npos,ntot-npos)~elev1+elev2+elev3+elev4+maxNDVI1+seNDVI
                  +Matern(1|longitude+latitude),HLmethod="HL(0,1)",data=Loaloa,
                  family=binomial(),ranFix=list(nu=0.5,rho=2.255197,lambda=1.075))   

 ## longer computation requiring interpolation of 197 points 
 if (requireNamespace("maps")) { ## required for add.map=TRUE 
  filled.mapMM(lfit,add.map=TRUE,plot.axes=quote({axis(1);axis(2)}),
             decorations=quote(points(pred[,coordinates],pch=15,cex=0.3)),
             plot.title=title(main="Inferred prevalence, North Cameroon",
                                xlab="longitude",ylab="latitude"))
  }
}

# }