Reciprocal scaling after a correspondence analysis

performs the canonical graph of a correspondence analysis.

Usage

# S3 method for coa
score(x, xax = 1, dotchart = FALSE, clab.r = 1, clab.c = 1, 
    csub = 1, cpoi = 1.5, cet = 1.5, ...)
reciprocal.coa(x)

Arguments

x: an object of class coa
xax: the column number for the used axis
dotchart: if TRUE the graph gives a "dual scaling", if FALSE a "reciprocal scaling"
clab.r: a character size for row labels
clab.c: a character size for column labels
csub: a character size for the sub-titles, used with par("cex")*csub
cpoi: a character size for the points
cet: a coefficient for the size of segments in standard deviation
...: further arguments passed to or from other methods

Value

return a data.frame with the scores, weights and factors of correspondences (non zero cells)

Details

In a "reciprocal scaling", the reference score is a numeric code centred and normalized of the non zero cells of the array which both maximizes the variance of means by row and by column. The bars are drawn with half the length of this standard deviation.

References

Thioulouse, J. and Chessel D. (1992) A method for reciprocal scaling of species tolerance and sample diversity. Ecology, 73, 670--680.

Author

Daniel Chessel

Examples

layout(matrix(c(1,1,2,3), 2, 2), resp = FALSE)
data(aviurba)
dd1 <- dudi.coa(aviurba$fau, scan = FALSE)
score(dd1, clab.r = 0, clab.c = 0.75)
recscal <- reciprocal.coa(dd1)
head(recscal)
#>             Scor1         Scor2 Row Col     Weight
#> R12Sp1 -1.4223264 -1.8505207282 R12 Sp1 0.00132626
#> R44Sp1 -2.2557912 -1.8885813170 R44 Sp1 0.00132626
#> R46Sp1 -1.2464842 -2.0492095800 R46 Sp1 0.00132626
#> R3Sp2  -1.5188080 -0.7740337809  R3 Sp2 0.00265252
#> R11Sp2 -0.5136887 -0.0006023528 R11 Sp2 0.00132626
#> R19Sp2 -0.2777474 -0.7370905113 R19 Sp2 0.00132626
abline(v = 1, lty = 2, lwd = 3)
sco.distri(dd1$l1[,1], aviurba$fau)
sco.distri(dd1$c1[,1], data.frame(t(aviurba$fau)))


# 1 reciprocal scaling correspondence score -> species amplitude + sample diversity
# 2 sample score -> averaging -> species amplitude
# 3 species score -> averaging -> sample diversity

layout(matrix(c(1,1,2,3), 2, 2), resp = FALSE)
data(rpjdl)
rpjdl1 <- dudi.coa(rpjdl$fau, scan = FALSE)
score(rpjdl1, clab.r = 0, clab.c = 0.75)
if (requireNamespace("MASS", quietly = TRUE)) {
   data(caith, package = "MASS")
   score(dudi.coa(caith, scan = FALSE), clab.r = 1.5, clab.c = 1.5, cpoi = 3)
   data(housetasks)
   score(dudi.coa(housetasks, scan = FALSE), clab.r = 1.25, clab.c = 1.25, 
        csub = 0, cpoi = 3)
}

par(mfrow = c(1,1))
score(rpjdl1, dotchart = TRUE, clab.r = 0)