ScatterPlot: Enhanced scatterplot with fitting lines
In ModelDataComp: Model-Data Comparison

Description Usage Arguments Details Value Author(s) See Also Examples

This function creates a scatterplot with fitting lines. Points and fitting lines can be also computed by groups.

ScatterPlot(x, y, groups = NULL, plot.type = NULL, objfct = FALSE, 
    which.objfct = c("Cor", "Pbias", "RMSE"), add = FALSE, density.levels = 50, 
    density.labels = FALSE, image.nbrks = 300, fits = c("spline"), 
    fit.quantile = NULL, fit.minnobs = 10, fit.groups = TRUE, 
    fit.global = TRUE, col.points = "black", col.image = NULL, 
    col.density = NULL, col.fit = col.points, col.global = NULL, 
    xlab = "", ylab = "", main = "", xlim = NULL, ylim = NULL, 
    lwd = 2, lty = 1, quantile.x = c(0.01, 0.99), quantile.y = c(0.01, 
        0.99), plot = TRUE, cex = NULL, pch = 21, alpha = NULL, 
    nrpoints = NULL, legend = TRUE, ...)

`x`	vector of x values
`y`	vector of y values
`groups`	vector of grouping variables (optional), in case of NULL all points are treated as one group
`plot.type`	plot type: possible arguments are 'density' to plot density lines, 'points' to plot points, or 'image' to plot point counts as a raster image. In case of NULL, the optimal plot.type will be determined dependent on the number of values. It is also possible to combine plot types (e.g. c("image", "density")).
`objfct`	Compute objective functions and add a 1:1 line to the plot?
`which.objfct`	Which objective functions should be added to legend of the plot if objfct=TRUE?
`add`	add plot to existing plot?
`density.levels`	number of levels for density lines
`density.labels`	add labels to density lines?
`image.nbrks`	minimum number of breaks for the image
`fits`	Fitting methods that should be used. See `MultiFit` for details. If several methods area provided the mean of all methods is plotted.
`fit.quantile`	Compute fit to a certain quantile? If NULL, fits to the mean. Otherwise sa value between 0 and 1 can be specified to fit to a certain quantile.
`fit.minnobs`	minimum number of observations per group in order to calculate a fitting curve
`fit.groups`	Plot fit lines for groups?
`fit.global`	Plot global fit lines?
`col.points`	color for the points/groups
`col.image`	color range for image
`col.density`	color range for density lines
`col.fit`	colors for fitting lines
`col.global`	colors for global fittiing line
`xlab`	a title for the x axis
`ylab`	a title for the y axis
`main`	an overall title for the plot
`xlim`	range for the x axis
`ylim`	range for the y axis
`lwd`	line width
`lty`	line type
`quantile.x`	lower and upper quantile to exclude extreme x values from plotting and fit calculations
`quantile.y`	lower and upper quantile to exclude extreme y values from plotting and fit calculations
`plot`	should a scatterplot be plotted or calculate only fitting lines
`cex`	size of the points in the plot. If NULL point size will be calculated based on the number of points.
`pch`	point symbol
`alpha`	transparency of the points (0...255)
`nrpoints`	number of outlier points to be drawn on images
`legend`	dsiplay a legend?
`...`	further arguments to `MultiFit` or `plot.default`

This function plots a scatterplot.

The function returns a list with the computed fitting lines and objective functions per group.

Matthias Forkel <matthias.forkel@geo.tuwien.ac.at> [aut, cre]

ObjFct

# create some data:
n <- 10000
x <- runif(n, 0, 100)
groups <- c(rep(1, 3000), rep(2, 3000), rep(3, 4000))
y <- (3 * x^(1/2) + rnorm(n, 0, x/20)) * groups

# standard plot: not very well distinguishable
plot(x, y)

# ScatterPlot 
result <- ScatterPlot(x, y)
result <- ScatterPlot(x, y, col.image=c("red", "grey", "blue"))

# ScatterPlot with coulored groups and fitting lines
result <- ScatterPlot(x, y, groups)

# different plot types 
result <- ScatterPlot(x, y, plot.type="points")
result <- ScatterPlot(x, y, plot.type="density")
result <- ScatterPlot(x, y, plot.type=c("image", "density"))
result <- ScatterPlot(x, y, groups, plot.type=c("image", "points"))

# plot and compute objective functions
result <- ScatterPlot(x, y, groups, objfct=TRUE)
result

# plot fits to upper 0.9 and lower 0.1 quantiles, mean fit from two methods
result <- ScatterPlot(x, y, groups, fits=c("poly3", "spline"), 
 fit.quantile=0.9, plot.type="image", fit.global=FALSE)
result <- ScatterPlot(x, y, groups, fits=c("poly3", "spline"), 
 fit.quantile=0.1, plot.type=NA, add=TRUE, fit.global=FALSE)