cplot {metRology}

Consistency plot for Key Comparisons

Description

Produces a consistency plot for typical metrology comparison data.

Usage

  cplot(x,u,labels=names(x), p.adjust.method="holm", ordered=TRUE,
        breaks=c(0,0.001,0.01, 0.05, 0.1,1), 
        col=terrain.colors(length(breaks)-1), log.p=FALSE,
        main=paste("Consistency map -", deparse(substitute(x))), 
        subtitle=NULL, key=FALSE, 
        key.width=2.54, key.height=0.6,...)

Arguments

x

Vector of reported values

u

Vector of length length(x) of standard uncertainties

labels

Vector of of length length(x) labels for x-axis marks.

p.adjust.method

p-value adjustment method; passed to p.adjust.

ordered

If TRUE (the default) observations are arranged in ascending order of x.

breaks

Vector of breaks; passed to image.

col

Vector of colours of length length(breaks)-1. Passed to image.

log.p

If TRUE, the plot shows -log_10(p).

main, subtitle

Main and subtitle for plot.

key

If TRUE a key is added to the plot.

key.width, key.height

Width and height of key, if plotted. See details for specification.

...

Graphical parameters passed to image. If present, cex.axis is passed to axis for main figure axes only.

Details

Calculates the (square, symmetric matrix of) optionally adjusted p-values for a two-tailed z-test of |x[i]-x[j]|/sqrt(u[i]^2+u[j]^2) against zero and plots the p-values as an image.

p.adjust is called prior to plotting to correct for multiple comparisons. To suppress adjustment, set p.adjust.method="none".

key.height is a fraction of the figure region height. key.width is the width of the key area in cm, unless under 1, in which case it is interpreted as a fraction of the plot region width.

If log.p is TRUE and subtitle NULL, a subtitle indicating the use of log.p is added to the plot,

Value

Invisibly returns a matrix of pairwise test p-values or, if log.p==TRUE, matrix of -log_10(p).

Author(s)

S Ellison [e-mail]

See Also

p.adjust, image.

Examples

  data(Pb)
  cplot(Pb$value, Pb$u, key=TRUE)