plot.mask: Plot Habitat Mask, Density or Resource Surface

In secr: Spatially Explicit Capture-Recapture

Plot Habitat Mask, Density or Resource Surface

Description

Plot a habitat mask either as points or as an image plot. Colours maybe used to show the value of one mask covariate.

Usage

## S3 method for class 'mask'
plot(x, border = 20, add = FALSE, covariate = NULL, axes = FALSE,
    dots = TRUE, col = "grey", breaks = 10, meshcol = NA, ppoly = TRUE,
    polycol = "red", legend = TRUE, ...)

## S3 method for class 'Dsurface'
plot(x, covariate, group = NULL, plottype =
    "shaded", scale = 1, ...)

## S3 method for class 'Rsurface'
plot(x, covariate = "Resource", plottype =
    "shaded", scale = 1, ...)

spotHeight (object, prefix = NULL, dec = 2, point = FALSE, text = TRUE,
    sep = ", ",  session = 1, scale = 1, ...)

Arguments

x, object	mask or Dsurface object
border	width of blank display border (metres)
add	logical for adding mask points to an existing plot
covariate	name (as character string in quotes) or column number of a covariate to use for colouring
axes	logical for plotting axes
dots	logical for plotting mask points as dots, rather than as square pixels
col	colour(s) to use for plotting
breaks	an integer or a numeric vector – see cut
meshcol	colour for pixel borders (NA for none)
ppoly	logical for whether the bounding polygon should be plotted (if ‘poly’ specified)
polycol	colour for outline of polygon (ppoly = TRUE)
legend	logical; if TRUE a legend is plotted
...	other arguments passed to eqscplot (in the case of plot.mask), plot.mask (in the case of plot.Dsurface and plot.Rsurface), and points or text (in the case of spotHeight)
group	group for which plot required, if more than 1
plottype	character string c("dots", "shaded", "contour", "persp")
scale	numeric multiplier for density or other numeric covariate (see Dsurface)
prefix	character vector for name(s) of covariate(s) to retrieve
dec	number of decimal places for rounding density
point	logical for whether to plot point
text	logical for whether to place density label on plot
sep	character separator for elements if length(prefix)>1
session	session number or identifier

Details

The argument dots of plot.mask selects between two distinct types of plot (dots and shaded (coloured) pixels).

plot.Dsurface and plot.Rsurface offer contour and perspective plots in addition to the options in plot.mask. It may take some experimentation to get what you want - see contour and persp.

For plot.Dsurface the default value of ‘covariate’ is ‘D’ unless the Dsurface has a ‘parameter’ attribute of ‘noneuc’,

If using a covariate or Dsurface or Rsurface to colour dots or pixels, the col argument should be a colour vector of length equal to the number of levels (the default palette from 2.9.0 is terrain.colors, and this palette will also be used whenever there are too few levels in the palette provided; see Notes for more on palettes). Border lines around pixels are drawn in ‘meshcol’. Set this to NA to eliminate pixel borders.

If a covariate is specified in a call to plot.Dsurface then that covariate will be plotted instead of density. This is a handy way to contour a covariate (contouring is not available in plot.mask).

If ‘breaks’ is an integer then the range of the covariate is divided into this number of equal intervals. Alternatively, ‘breaks’ may be a vector of break points (length one more than the number of intervals). This gives more control and often ‘prettier’

spotHeight may be used to interrogate a plot produced with plot.Dsurface or plot.Rsurface, or by plot.mask if the mask has covariates. prefix defaults to ‘density.’ for Dsurface objects and to '' (all covariates) for mask objects. The predicted density or covariate at the nearest point is returned when the user clicks on the plot. Multiple values may be displayed (e.g., prefix = c("lcl","ucl") if Dsurface includes confidence limits). Click outside the mask or hit the Esc key to end. spotHeight deals with one session at a time.

Legend plotting is enabled only when a covariate is specified. It uses legend when dots = TRUE and strip.legend otherwise.

Value

If covariate is specified and plottype = "shaded" then plot.mask invisibly returns a character vector of the intervals defined by ‘breaks’ (useful for plotting a legend).

If plottype = "persp" then plot.mask invisibly returns a the perspective matrix that may be used to add to the plot with trans3d.

spotHeight invisibly returns a dataframe of the extracted values and their coordinates.

Note

plot.mask() acquired the argument ‘legend’ in version 2.9.0, and other changes (e.g., breaks = 10) may alter the output.

Contouring requires a rectangular grid; if a Dsurface is not rectangular then plot.Dsurface with plottype = "contour" triggers a call to rectangularMask.

The colour palettes topo.colors, heat.colors and terrain.colors may be viewed with the demo.pal function in the Examples code of their help page palettes.

The package RColorBrewer is a good source of palettes. Try display.brewer.all() and e.g., col = brewer.pal(7, "YlGn").

See Also

colours, mask, Dsurface, rectangularMask, contour persp strip.legend

Examples

# simple

temptrap <- make.grid()
tempmask <- make.mask(temptrap)
plot (tempmask)

## Not run: 

## restrict to points over an arbitrary detection threshold,
## add covariate, plot image and overlay traps

tempmask <- subset(tempmask, pdot(tempmask, temptrap,
    noccasions = 5)>0.001)
covariates (tempmask) <- data.frame(circle = 
    exp(-(tempmask$x^2 + tempmask$y^2)/10000) )
plot (tempmask, covariate = "circle", dots = FALSE, axes = TRUE, 
    add = TRUE, breaks = 8, col = terrain.colors(8), mesh = NA)
plot (temptrap, add = TRUE)

## add a legend
par(cex = 0.9)
covrange <- range(covariates(tempmask)$circle)
step <- diff(covrange)/8
colourlev <- terrain.colors(9)
zlev <- format(round(seq(covrange[1],covrange[2],step),2))
legend (x = "topright", fill = colourlev, legend = zlev, 
    y.intersp = 0.8, title = "Covariate")

title("Colour mask points with p.(X) > 0.001")
mtext(side=3,line=-1, "g0 = 0.2, sigma = 20, nocc = 5")

## Waitarere possum density surface extrapolated across region

regionmask <- make.mask(traps(possumCH), buffer = 1000, spacing = 10,
    poly = possumremovalarea)
dts <- distancetotrap(regionmask, possumarea)
covariates(regionmask) <- data.frame(d.to.shore = dts)
shorePossums <- predictDsurface(possum.model.Ds, regionmask)

## plot as coloured pixels with white lines
colourlev <- terrain.colors(7)
plot(shorePossums, breaks = seq(0,3.5,0.5), plottype = "shaded",
    poly = FALSE, col = colourlev, mesh = NA)
plot(traps(possumCH), add = TRUE, detpar = list(col = "black"))
polygon(possumremovalarea)

## check some point densities
spotHeight(shorePossums, dec = 1, col = "black")

## add a legend
zlev <- format(seq(0,3,0.5), digits = 1)
legend (x = "topright", fill = colourlev, legend =
    paste(zlev,"--"), y.intersp = 1, title = "Density / ha")

## End(Not run)

secr documentation built on Nov. 4, 2024, 9:06 a.m.