* data cleaning (if you have a list of data sets that could/should be
  tweaked, that would be great, but we should maybe go through them
  all systematically anyway)

* moving some code to the mosaic package and making mosaic a
  requirement for abd (simplifies code maintenance -- I've already
  begun this with my fastR package)

* adding more demos/examples

rda to fix
    Eelgrass (?)
    LizardSprint (?)

Use rptR or ICC for repeatability.

Implement a manual viewer:
abd.manual <- function(){
  viewer <- getOption("pdfviewer")
  file <- system.file("doc/abd-manual.pdf", package = "abd")
  system(paste(viewer, file, sep = " "))
  invisible()
}

Fig. 2.4-2
Fig. 4.3-1
Fig. 8.1-1 ggplot

Functions
  sign.test()
  Wald adjusted prop.ci
  Use odds.ratio() from HH; write print method

Error Checking
  sum.of.products()
  sum.of.squares()
  se() to sem()

