ellipsescatter {LSD}

Visualize subgroups of two-dimensional data assuming normal distributions

Description

A scatterplot with additional colored ellipses based on a gaussianity assumption.

Usage

ellipsescatter(
  x,
  y,
  groups,
  colors = NULL,
  pch = 20,
  bgcol = "darkgrey",
  main = "ellipsescatter",
  xlab = NULL,
  ylab = NULL,
  scalesd = 1,
  level = 0.75,
  legend.cex = 1,
  location = "topright",
  ...
)

Arguments

x

a numeric vector.

y

a numeric vector.

groups

a list of indices or vector names to be plotted as groups (not necessarily all of x and y).

colors

a character vector of R build-in colors corresponding to the chosen groups.

pch

the plotting character (to be passed to plot).

bgcol

a R build-in color for non-grouped points.

main

title(s) of the plot, standard graphics parameter.

xlab

x label, standard graphics parameter.

ylab

y label, standard graphics parameter.

scalesd

a numeric value giving the scaling factor for standard deviations in each dimension (defaults to 1).

level

a numeric value (between 0 and 1) giving the confidence level of a pairwise confidence region.

legend.cex

a numerical value giving the amount by which the added legend should be magnified relative to the default.

location

the x and y co-ordinates to be used to position the legend (see 'xy.coords').

...

additional parameters to be passed to points and plot.

Author(s)

Bjoern Schwalb

See Also

demotour, disco, colorpalette

Examples

x = c(rnorm(50),rnorm(100,2),rnorm(50,4))
y = (x + rnorm(200,0,0.8))*rep(c(1,4,1),c(50,100,50))
x = sign(x)*abs(x)^1.3

groups = list("Green" = 1:50,"Red" = 51:150,"Blue" = 151:200)
colors = c("darkgreen","darkred","darkblue")
ellipsescatter(x,y,groups,colors,location = "topleft")