• Open-source programming language for data analytics and graphic representation
• Input from common data file types: CSV, Excel, SQL
• Outputs to common graphic types: SVG, PNG, JPG, PDF
• Report to common document types: PDF, HTML, LaTeX, XML
• Supported by leading professional and research societies for mathematics, statistics and informatics
• Well suited for both sampled and Big Data analysis

• Text based scripting in interactive mode
• Script execution console displays textual results
• Graphics output canvas

• SPSS-like chooser of graphic and statistical procedures
• Excel-like data editor
• Factor and analysis selection translated to R code
• The chooser translates the parameters into valid R commands

• Full support for documentation
• Cloud based publishing of presentations and reports
• Supports interactive statistical programming

# n = 10000
plot(x,y)

# n = 10000
smoothScatter(x, y)

data = rbind(
c("id","math","engl",
   "biol","music","R-code"
c(MAX,  20, 20, 20, 20, 20),
c(MIN,   0,  0,  0,  0,  0),
c(mr-a,  3,  9,  5, 20, 15),
c(mr-b, 19,  4, 12,  6,  2),
c(mr-c,  9, 15, 16,  2, 18))

radarchart(data, axistype=1, 
  caxislabels=
    seq(0,20,5), 
  vlcex=0.8)

Page User Data Matrix

     Rg Fn Lb Bu Rs Hm
guest 0  0  1  0  1  5
appl  4  2  0  0  0  2
stud  8  4  2  0  0  2
staff 1  6  2  4  1  3
instr 8  4  4  2  4  2
admin 4  6  1  6  2  3

chordDiagram(pgUsrDatMx)

data        15
statistics   5
probability  3  ...

library(wordcloud2)

wordcloud2(wrddat,
  color=ifelse(
    wrddat[,2] > 13,
      'red','skyblue'))

NYTimes Shiny Server URL

require(maps)
library(mapdata)
countries = 
  c('thailand','malaysia',
  'indonesia','philippines',
  'singapore','myanmar',
  'laos','vietnam',
  'cambodia')
colors = rainbow(128)
map('world',countries,
  fill=TRUE,col=colors)

Provides different political boundary maps

• National
• Changwat
• Amphur
• Tambon

Choose specific columns and checked for correlation

cardat =
 cars[,c(1,3:6)]
pairs(cardat)

library(corrgram)
corrgram(cardat, order=TRUE, 
 lower.panel=panel.shade,
 upper.panel=panel.pie,
 text.panel=panel.txt,
 main="Car Milage Data")

Reorders the dataset to minimize the differences between neighboring cars and attributes

library(rpart)
pfit = rpart(mpg ~ wt + hp + drat + disp,data=cardat)
summary(pfit)
plot(pfit,compress=TRUE)
text(pfit,use.n=TRUE)

Shiny URL

model = HoltWinters(tourist)
p = predict(model,12,
  prediction.interval=
  TRUE,level=0.80)

m = arima(tourist,
  order = c(3,1,1), 
  seasonal = list(
    order=c(2,1,1),
    period=4))
pp = predict(m,12)

A prize is hidden behind one door

The Rules:

• A contestant selects 1 of 3 doors.
• The announcer shows a lesser prize behind another door
• The contestant can keep his choice or choose the remaining door.

Which door should the contestant choose?

Make a list of 10,000 preferred order for selecting the three doors.

R-code:

dat = rep(sample(3),10000)
dim(dat) = c(10000,3)
colnames(dat) =
  c("c1","c2","c3")

First 5 entries:

     c1 c2 c3
[1,]  3  1  2
[2,]  1  2  3
[3,]  2  3  1
[4,]  3  1  2
[5,]  1  2  3

tally = rep(0,9)
dim(tally) = c(3,3)
for (i in 1:3) {
 for (j in 1:3) {
  tally[i,j] = 
   sum(dat[,i] == j)}}
barplot(tally, col=c(
  "red","brown","green"),
  names.arg = 1:3,
  xlab="Preferences",
  main="Tally of Doors")

R-code:

prizeloc =
 ceiling(runif(10000,0,3))

First 20 entries:

 [1] 3 2 3 1 1 1 1 1 2 2 1 2 3 1 3 2 1 1 2 3

Tally the number of times that the winning door was selected either as first choice or alternative choice.

wins = c(
sum(dat[,1]==prizeloc),
sum(dat[,2]==prizeloc) +
  sum(dat[,3]==prizeloc))