CLME (version 1.0-1)

plot.clme: S3 method to plot objects of class clme

Description

Generates a basic plot of estimated $\theta$ coefficients which are subject to constraints. Lines indicate individual constraints (not global tests) and significance.

Usage

## S3 method for class 'clme':
plot( x , alpha=0.05 , place.leg="below" , inset=0.01,
            ci.wd=0, plot.names=TRUE, ylim, cex=1.75, pch=21, bg="white" ,  
            xlab = expression(paste("Component of ", theta[1])),
            ylab = expression(paste("Estimated Value of " , theta[1])), 
           ...)

Arguments

x

object of class 'clme'.

alpha

significance level of the test.

place.leg

character indicating placement of legend. See Details.

inset

inset distance(s) from the margins as a fraction of the plot region when legend is placed by keyword.

ci.wd

width of bars for the confidence limits.

plot.names

logical, whether to plot the group names on the X-axis.

ylim

limits of the y axis.

cex

size of plotting symbols.

pch

plotting symbols.

bg

background (fill) color of the plotting symbols.

xlab

label of the x axis.

ylab

label of the y axis.

...

additional plotting arguments.

Details

All of the individual contrasts in the constraints$A matrix are tested and plotted. The global contrasts from constraints$B are not shown (unless they happen to align with the individual contrasts). Only the elements of $\theta$ which appear in any constraints (e.g. the elements of $\theta_{1}$) are plotted. Coefficients for the covariates are not plotted. Solid lines denote no significant difference, while dashed lines denote statistical significance. Significance is determined by the individual p-value being less than or equal to the supplied $\alpha$ threshold. By default a legend denoting the meaning of solid and dashed lines will be placed below the graph. Argument place.leg may be set to a legend keyword (e.g. legend=''bottomright'') to place it inside the graph at the specified location. Setting place.leg to FALSE or to a non-supported keyword suppresses the legend. Confidence intervals for the coefficients may be plotted, but the user is cautioned about these. They are individual confidence intervals, and are computed using the covariance matrix of the unconstrained estimates of $\theta_{1}$. As a result, the confidence intervals will be wider and sometimes appear to conflict with the significance tests.

See Also

CLME-package, constrained.lme

Examples

set.seed( 42 )

n  <- 5
P1 <- 5

X1 <- diag(P1) %x% rep(1,n)
X2 <- as.matrix( rep(1,P1) %x% runif(n , 0,2) )
U  <- rep(1,P1) %x% diag(n)
X  <- as.matrix( cbind(X1,X2) )

tsq <- 1
ssq <- 0.7

Nks <- dim(X1)[1]
Qs  <- dim(U)[2]

xi <- rnorm( sum(Qs)  , 0 , rep(sqrt(tsq) , Qs)  )
ep <- rnorm( sum(Nks) , 0 , rep(sqrt(ssq) , Nks) )  

thetas <- c(2 , 3 , 3, 3 , 4 , 2 )
Y      <- X%*%thetas + U%*%xi + ep
const  <- list(order='simple' , decreasing=FALSE)

# Note that 'nsim' has been reduced for illustration

clme.out <- constrained.lme( method='PAVA', Y=Y, X1=X1, X2=X2, U=U,
                     constraints=const , nsim=10 )

plot.clme( clme.out )