compositions (version 2.0-4)

colorsForOutliers: Create a color/char palette or for groups of outliers

Description

Conveniance Functions to generate meaningfull color palettes for factors representing different types of outliers.

Usage

colorsForOutliers1(outfac, family=rainbow,
                          extreme="cyan",outlier="red",ok="gray40",unknown="blue")
colorsForOutliers2(outfac,use=whichBits(gsi.orSum(levels(outfac))),
                        codes=c(2^outer(c(24,16,8),1:7,"-")),ok="yellow")
pchForOutliers1(outfac,ok='.',outlier='\004',extreme='\003',unknown='\004',...,
  other=c('\001','\002','\026','\027','\010','\011','\012','\013','\014','\015',
    '\016',strsplit("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ","")[[1]])
    )

Value

a character vector of colors or a numeric vector of plot chars.

Arguments

outfac

a factor given by an OutlierClassifier (e.g. OutlierClassifier1). colorsForOutliers1 is used for the types "best","type","outlier","grade". colorsForOutliers2 is used for type all.

family

a function generating a color palette from a numer of colors requested.

extreme

The color/char for extrem but not definitivly outlying observations.

outlier

The color/char for detected outliers.

unknown

The color/char for observation with unclear classification.

other

The character codes for other outlier classes.

ok

The color/char for nonoutlying usual observations.

use

a numerical vector giving the indices of the bits of the output to be represented. The sequence of the bits determins how each bit is represented.

codes

The color influences to be used for each bit.

...

further codings for other factorlevels

Author

K.Gerald v.d. Boogaart http://www.stat.boogaart.de

Details

This functions are provided for coveniance to quickly generate a palette of reasonable colors or plotting chars for groups of outliers classified by OutlierClassifier1.

See Also

OutlierClassifier1

Examples

if (FALSE) {
data(SimulatedAmounts)
data5 <- acomp(sa.outliers5)
olc <- OutlierClassifier1(data5)
plot(data5,col=colorsForOutliers1(olc)[olc])
olc <- OutlierClassifier1(data5,type="all")
plot(data5,col=colorsForOutliers2(olc)[olc])
}