plot.MCRestimate: Plot method for a objects of class MCRestimate

In MCRestimate: Misclassification error estimation with cross-validation

Description

plot.MCRestimate visualizes a 'vote matrix'. A 'vote matrix' is the result of a classification procedure. For every sample (=row) i and every class (=column) j the matrix element [i,j] is the probability or frequency the classification method predicts sample i as a member of class j.

Usage

## S3 method for class 'MCRestimate'
plot(x,
                 class.factor=NULL,
                 rownames.from.object=FALSE,
                 sample.order=TRUE,
                 legend=FALSE,
                 mypalette=NULL,
                 shading=NULL,
                 xlab="Sample ID",
                 ylab="Frequency of correct classification",
                 cex.axis=1,...)

Arguments

x	Object of S3 class MCRestimate or a matrix
class.factor	Factor. Its length must match the number of rows in x and the levels must be the same as the colnames in x. If x is of class MCRestimate this argument will be ignored.
rownames.from.object	Logical. If TRUE then the rownames of the matrix or the sample names of MCRestimate in x are used as labels for the x-axis
sample.order	Logical. If TRUE then the samples are ordered by class membership
legend	Logical. If TRUE then there will be a small legend in the output
mypalette	vector with length equal to the number of classes. The vector specifies the color for the bar representing the classes. If 'NULL' colors chosen by the author are used.
shading	the density of shading lines for the rectangles that indicate the groups, in lines per inch. The default value of 'NULL' means that no shading lines are drawn.
xlab	Character
ylab	Character
cex.axis	numeric
...	Further arguments that are passed on to plot.default

Value

The function is called for its side effect, creating a plot on the active graphics device.

Author(s)

Markus Ruschhaupt mailto:[e-mail]

See Also

MCRestimate

Examples

  x <- c(0.5, 0.3, 0.7, 0.3, 0.8, 0.2, 0.3)
  mat2 <- cbind(x, 1-x)
  fac2 <- factor(c("low", rep("high", 3), rep("low", 3)))
  colnames(mat2) <- levels(fac2)

  mat3 <- cbind(x/3, 2*x/3, 1-x)
  fac3 <- factor(c(rep("high", 3), rep("intermediate", 2), rep("low", 2)))
  colnames(mat3) <- levels(fac3)
if (interactive()) {
  x11(width=9, height=9)
  par(mfrow=c(3,1))}
  plot.MCRestimate(mat2, fac2)
  plot.MCRestimate(mat2, fac2, sample.order=FALSE)
  plot.MCRestimate(mat3, fac3)

MCRestimate documentation built on Oct. 31, 2019, 10:29 a.m.