reduce.cex.lr: reduce cex lr

Description

If left or right edges of the text are going out of the plotting region, then decrease cex until it fits. We call calc.boxes inside, so you should set cex before using this.

Usage

1

Arguments

d
...

Author(s)

Toby Dylan Hocking

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
if(require(ElemStatLearn) && require(lars) && require(ggplot2)){
  pros <- subset(prostate,select=-train,train==TRUE)
  ycol <- which(names(pros)=="lpsa")
  x <- as.matrix(pros[-ycol])
  y <- pros[[ycol]]
  fit <- lars(x,y,type="lasso")
  beta <- scale(coef(fit),FALSE,1/fit$normx)
  arclength <- rowSums(abs(beta))

  path.list <- list()
  for(variable in colnames(beta)){
    standardized.coef <- beta[, variable]
    path.list[[variable]] <-
      data.frame(step=seq_along(standardized.coef),
                 arclength,
                 variable,
                 standardized.coef)
  }
  path <- do.call(rbind, path.list)
  p <- ggplot(path,aes(arclength,standardized.coef,colour=variable))+
    geom_line(aes(group=variable))

  ## the legend isn't very helpful.
  print(p)

  ## add direct labels at the end of the lines.
  direct.label(p, "last.points")

  ## on my screen, some of the labels go off the end, so we can use
  ## this Positioning Method to reduce the text size until the labels
  ## are on the plot.
  direct.label(p, list("last.points","reduce.cex.lr"))

  ## the default direct labels for lineplots are similar.
  direct.label(p)
}

Questions? Problems? Suggestions? or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.