fadeRug: Fade out the areas in a surface without data.

In itsadug: Interpreting Time Series and Autocorrelated Data Using GAMMs

Fade out the areas in a surface without data.

Description

Add a transparency Rug to a contour plot or image.

Usage

fadeRug(
  x,
  y,
  n.grid = 30,
  too.far = 0.03,
  col = "white",
  alpha = 1,
  use.data.range = TRUE
)

Arguments

x	Observations on x-axis.
y	Observations on y-axis.
n.grid	Resolution of Rug. Defaults to 30, which means that the x- and y-axis are divided in 30 bins. A two-value vector ould be used to specify different bins for x- and y-axis.
too.far	plot grid nodes that are too far from the points defined by the variables given in view can be excluded from the plot. too.far determines what is too far. The grid is scaled into the unit square along with the view variables and then grid nodes more than too.far from the predictor variables are excluded. Based on exclude.too.far of Simon N. Wood.
col	Color representing missing data. Defaults to 'white'.
alpha	Transparency, number between 0 (completely transparent) and 1 (non-transparent). Defaults to 1.
use.data.range	Logical value, indicating whether x and y are the data that the plot is based on. Defaults to TRUE.

Value

Plots a shaded image over the contour plot or image.

Warning

On Linux x11 devices may not support transparency. In that case, a solution might be to write the plots immediately to a file using functions such as pdf, or png.

Author(s)

Jacolien van Rij, based on Simon N. Wood's exclude.too.far

See Also

rug, contour, image

Other Functions for plotting: rug_model()

Examples

data(simdat)

# Introduce extreme values:
set.seed(123)
newdat <- simdat[sample(which(simdat$Time < 1500),
    size=round(.5*length(which(simdat$Time < 1500)))),]
newdat <- rbind(newdat, 
    simdat[sample(which(simdat$Time > 1500),
    size=5),])
# Some simple GAM with tensor:
m1 <- bam(Y ~ te(Time, Trial), data=newdat)
# plot summed effects:
fvisgam(m1, view=c('Time', 'Trial'), zlim=c(-15,15))
fadeRug(newdat$Time, newdat$Trial)
# check with data points:
points(newdat$Time, newdat$Trial, pch=16, col=alpha(1))

# compare with default rug:
fvisgam(m1, view=c('Time', 'Trial'), zlim=c(-15,15))
rug(newdat$Time)
rug(newdat$Trial, side=2)
fadeRug(newdat$Time, newdat$Trial)
# and compare with too.far:
fvisgam(m1, view=c('Time', 'Trial'), zlim=c(-15,15),
    too.far=.03)
vis.gam(m1, view=c('Time', 'Trial'), zlim=c(-15,15),
    too.far=.03, plot.type='contour', color='topo')

# in case fade rug overlaps with color legend:
fvisgam(m1, view=c('Time', 'Trial'), zlim=c(-15,15),
     add.color.legend=FALSE)
fadeRug(newdat$Time, newdat$Trial, alpha=.75)
gradientLegend(c(-15,15), pos=.875)

# change x- and y-grid, and color:
fvisgam(m1, view=c('Time', 'Trial'), zlim=c(-15,15))
points(newdat$Time, newdat$Trial)
fadeRug(newdat$Time, newdat$Trial, n.grid=c(100,10), col='gray')

itsadug documentation built on June 17, 2022, 5:05 p.m.