plot function

Plot Diagnostics for Multiscale Bootstrap

plot method for class "scaleboot".

## S3 method for class 'scaleboot'
plot(x, models=NULL, select=NULL, sort.by=c("aic","none"),
     k=NULL, s=NULL, sp=NULL, lambda=NULL, bpk=NULL,
     xval = c("square", "inverse","sigma"),
     yval = c("psi", "zvalue", "pvalue"), xlab = NULL,
     ylab = NULL,log.xy = "", xlim = NULL, ylim = NULL,
     add = F, length.x = 300, main=NULL,
     col =1:6, lty = 1:5, lwd = par("lwd"), ex.pch=2:7,
     pch = 1, cex = 1, pt.col = col[1],pt.lwd = lwd[1],
     legend.x = NULL, inset = 0.1, cex.legend=1,...)

## S3 method for class 'summary.scaleboot'
plot(x, select="average",
     k=x$parex$k,s=x$parex$s,sp=x$parex$sp,lambda=x$parex$lambda, ...)

## S3 method for class 'scalebootv'
plot(x,models=attr(x,"models"),sort.by="none",...)

## S3 method for class 'summary.scalebootv'
plot(x, select="average",...)

## S3 method for class 'scaleboot'
lines(x,z,models=names(x$fi), k=NULL,s=NULL,sp=NULL,lambda=NULL,
      bpk=NULL, length.x=z$length.x, col=z$col,lty=z$lty,lwd=z$lwd,... )

sblegend(x="topright",y=NULL,z,inset=0.1,...)

sbplotbeta(beta, p=0.05, col.contour=c("blue","red","green"),
                     drawcontours = TRUE, drawlabels = TRUE,
                     labcex=1,length=100, cex=1, col="black",
                     xlim=NULL, ylim=NULL, lim.countourexpand=0 )

Arguments

• x: an object used to select a method. For sblegend, x is a numeric or character such as "lefttop" or "righttop", which is passed to legend.
• models: character vector of model names. Numeric is also allowed.
• select: "average", "best", or one of the fitted models.
• sort.by: "aic" or "none".
• k: k for extrapolation.
• s: s for extrapolation.
• sp: sp for extrapolation.
• lambda: a numeric of specifying the type of p-values; Bayesian (lambda=0) Frequentist (lambda=1).
• bpk: (experimental for 2-step bootstrap)
• xval: specifies x-axis. "square" for $\sigma^2$, "inverse" for $1/\sigma$, "sigma" for $\sigma$.
• yval: specifies y-axis. "zvalue" for $\psi(\sigma^2|\beta)/\sigma$ or qnorm(1-bp[i]), "pvalue" for $1-\Phi(\psi(\sigma^2|\beta)/\sigma)$ or bp[i], "psi" for $\psi(\sigma^2|\beta)$ or sqrt(sa[i])*qnorm(1-bp[i]).
• xlab: label for x-axis.
• ylab: label for y-axis.
• log.xy: character to specify log-scale. "", "x", "y", or "xy".
• xlim: range for x-axis.
• ylim: range for y-axis.
• add: logical for adding another plot.
• length.x: the number of segments to draw curves.
• main: for title.
• col: color for model curves.
• lty: lty for model curves.
• lwd: lwd for model curves.
• ex.pch: pch for extrapolation.
• pch: pch for bp points.
• cex: cex for bp points.
• pt.col: col for bp points.
• pt.lwd: lwd for bp points.
• legend.x: passed to sblegend as the first argument.
• ...: further arguments passed to or from other methods.
• z: output from previous plot.scaleboot.
• y: numeric passed to legend.
• inset: inset distance from the margins, which is passed to legend.
• cex.legend: cex for legend
• beta: matrix of beta values. beta[,1] is beta0, beta[,2] is beta1.
• p: significance level for drawing contour lines.
• col.contour: colors for SI, AU, BP.
• drawcontours: draw contours when TRUE.
• drawlabels: draw labels at contours when TRUE.
• labcex: cex for contours.
• length: grid size for drawing contours.
• lim.countourexpand: expand contour plotting region

Details

The plot method plots bootstrap probabilities and calls the lines

method, which draws fitted curves for models.

Author(s)

Hidetoshi Shimodaira

See Also

sbfit, sbphylo.

Examples

data(mam15)
## a single plot
a <- mam15.relltest[["t4"]] # an object of class "scaleboot"
plot(a,legend="topleft") # x=sigma^2, y=psi
plot(a,xval="inverse",yval="zvalue",
     legend="topleft") # x=1/sigma, y=z-value
plot(a,xval="sigma",log="x",yval="pvalue",
     legend="topleft") # x=log(sigma), y=probability
## plot of extrapolation
plot(summary(a),legend="topleft")
## multiple plots
b <- mam15.relltest[1:15] # an object of class "scalebootv"
plot(b) # x=sigma^2, y=psi