plotsummary: graphical summaries of variables of a data set

In aplpack: Another Plot Package: 'Bagplots', 'Iconplots', 'Summaryplots', Slider Functions and Others

Description

plotsummary shows some important characteristics of the variables of a data set. For each variable a plot is computed consisting of a barplot, an ecdf, a density trace and a boxplot.

Usage

 
 plotsummary(data, trim = 0, types = c("stripes", "ecdf", "density", "boxplot"),
             y.sizes = 4:1, design = "chessboard", main, mycols = "RB")
 

Arguments

data	Data set for computing a graphical summary.
trim	trim defines the fraction of observation for trimming on both ends of the data.
types	vector of types of representation of the data set. The elements of the vector will induce small plots which are stacked in vertical order. The first letter of the types is sufficient for defining a type.
y.sizes	defines the relative sizes of the small plots. The values are divided by their sum to get percentages.
design	if design is chessboard the graphics device is fragmented into rows and cols. Otherwise the images of a variable build vertical stripes.
main	defines a title for the graphics.
mycols	allows to define some colors for the showing the regions separated by the quartils.

Details

plotsummary can be use for a quick and dirty inspection of a data matrix or a list of variables. Without further specification some representation of each of the variables is built and stacked into a plot. The sizes of the types of representation can be set as well as the layout design of the graphics device. It is helpful to trim the data before processing because outliers will often hide the interesting characteristics.

Author(s)

Peter Wolf, [e-mail]

See Also

pairs, summary, str

Examples

                                                        
 ##---- Should be DIRECTLY executable !! ----         
 ##-- ==>  Define data, use random,                      
 ##--\tor do  help(data=index)  for the standard data sets.   
 plotsummary(cars)
 plotsummary(cars, types=c("ecdf", "density", "boxplot"), 
             y.sizes = c(1,1,1), design ="stripes")
 plotsummary(c(list(rivers=rivers, co2=co2), cars), y.sizes=c(10,3,3,1), mycols=3)
 plotsummary(cars, design="chessboard")
 # find all matrices in your R
 ds.of.R <- function(type="vector"){
   dat <- ls(pos=grep("datasets",search()))
   dat.type <- unlist(lapply(dat,function(x) {       
      num <- mode(x<-eval(parse(text=x)))
      num <- ifelse(is.array(x),"array",num)
      num <- ifelse(is.list(x),"list",num)
      num <- ifelse(is.matrix(x),"matrix",num)
      num <- ifelse(is.data.frame(x),"matrix",num)
      num <- ifelse(num=="numeric","vector",num)
      num }))
   return(dat[dat.type==type])
 }
 namelist <- ds.of.R("matrix")
 # inspect the matrices one after the other
 for(i in seq(along=namelist)){
   print(i); print(namelist[i])
   xy <- get(namelist[i])
   # plotsummary(xy,y.sizes=4:1,trim=.05,main=namelist[i]) 
   # Sys.sleep(1)
 }
 

aplpack documentation built on Sept. 30, 2021, 5:08 p.m.