This is my influence plot function. It uses ggplot to visualize leverage and studentized errors in a balloon plot where the balloon size is scaled by Cook's distance.

plotInfluence.R

plotInfluence <- function (model, fill="white", 
                           outline="black", size=30) {
  require(ggplot2)
  if(!inherits(model, "lm")) 
    stop("You need to supply an lm object.")
  df<-data.frame(Residual=rstudent(model), 
                 Leverage=hatvalues(model), 
                 Cooks=cooks.distance(model), 
                 Observation=names(hatvalues(model)), 
                 stringsAsFactors=FALSE)
  myxint<-c(2*mean(df$Leverage), 3*mean(df$Leverage))
  inds<-intersect(which(abs(df$Residual) < 2), 
                  which( df$Leverage < myxint[1]))
  if(length(inds) > 0) df$Observation[inds]<-""
  ggplot(df, aes_string(x='Leverage', y='Residual', 
                        size='Cooks', label='Observation'), 
         legend=FALSE) +
    geom_point(colour=outline, fill=fill, shape=21) + 
    scale_size_area(max_size=size) + 
    theme_bw(base_size=16) + geom_text(size=4) + 
    geom_hline(yintercept=c(2,-2), linetype="dashed") + 
    geom_vline(xintercept=myxint, linetype="dashed") + 
    ylab("Studentized Residuals") + 
    xlab("Hat-Values") + labs(size="Cook's distance")
}