plot.MCmcmc {MethComp}

Plot estimated conversion lines and formulae.

Description

Plots the pairwise conversion formulae between methods from a MCmcmc object.

Usage

## S3 method for class 'MCmcmc'
plot(
  x,
  axlim = range(attr(x, "data")$y, na.rm = TRUE),
  wh.cmp,
  lwd.line = c(3, 1),
  col.line = rep("black", 2),
  lty.line = rep(1, 2),
  eqn = TRUE,
  digits = 2,
  grid = FALSE,
  col.grid = gray(0.8),
  points = FALSE,
  col.pts = "black",
  pch.pts = 16,
  cex.pts = 0.8,
  ...
)

Arguments

x

A MCmcmc object

axlim

The limits for the axes in the panels

wh.cmp

Numeric vector or vector of method names. Which of the methods should be included in the plot?

lwd.line

Numerical vector of length 2. The width of the conversion line and the prediction limits. If the second values is 0, no prediction limits are drawn.

col.line

Numerical vector of length 2. The color of the conversion line and the prediction limits.

lty.line

Numerical vector of length 2. The line types of the conversion line and the prediction limits.

eqn

Should the conversion equations be printed on the plot?. Defaults to TRUE.

digits

How many digits after the decimal point shoudl be used when printing the conversion equations.

grid

Should a grid be drawn? If a numerical vector is given, the grid is drawn at those values.

col.grid

What color should the grid have?

points

Logical or character. Should the points be plotted. If TRUE or "repl" paired values of single replicates are plotted. If "perm", replicates are randomly permuted within (item, method) befor plotting. If "mean", means across replicates within item, method are formed and plotted.

col.pts

What color should the observation have.

pch.pts

What plotting symbol should be used.

cex.pts

What scaling should be used for the plot symbols.

...

Parameters to pass on. Currently not used.

Value

Nothing. The lower part of a (M-1) by (M-1) matrix of plots is drawn, showing the pairwise conversion lines. In the corners of each is given the two conversion equations together with the prediction standard error.

See Also

MCmcmc, print.MCmcmc

Examples

## Not run: data( hba1c )
## Not run: str( hba1c )
## Not run: hba1c <- transform( subset( hba1c, type=="Ven" ),
                    meth = dev,
                    repl = d.ana )
## End(Not run)
## Not run: hb.res <- MCmcmc( hba1c, n.iter=50 )
## Not run: data( hba.MC )
## Not run: str( hba.MC )
## Not run: par( ask=TRUE )
## Not run: plot( hba.MC )
## Not run: plot( hba.MC, pl.obs=TRUE )