R Programming 101

• R is a free programming language and software environment for statistical computing and graphics
• Developed by Ross Ihaka and Robert Gentleman at the University of Auckland, New Zealand in 1993.

• One of the most widely used statistical software programs
• FREE!
• Many online resources available
• Comprehensive R library and powerful visualization tools
• all-in-one functionality (presentations, reports, interactive web applications, etc.)

  #Step 1: Flip coin ten times
  num=sample(0:1,10,replace=TRUE)
  num
  #Step 2: Total number of heads
  total=sum(num)
  total
  #Step 3: Compute proportion
  prop=mean(num)
  prop

props=c()
  #Step 0: The number of students
n=25
for (i in 1:n){ 
  #Step 1: Each student flips coin ten times
  num=sample(0:1,10,replace=TRUE)
  #Step 2: Total number of heads per student
  total=sum(num)
  #Step 3: Compute proportion
  prop=mean(num)
  #Step 4: Combine all students' results
  props=append(props,prop) 
}
mean(props)

[1] 0.464

hist(props)

Does your histogram look normal?

props=c()
  #Step 0: The number of students
n=100
for (i in 1:n){ 
  #Step 1: Each student flips coin ten times
  num=sample(0:1,10,replace=TRUE)
  #Step 2: Total number of heads per student
  total=sum(num)
  #Step 3: Compute proportion
  prop=mean(num)
  #Step 4: Combine all students' results
  props=append(props,prop) 
}
mean(props)

[1] 0.497

hist(props)

Does your histogram look normal?

props=c()
  #Step 0: The number of students
n=2000
for (i in 1:n){ 
  #Step 1: Each student flips coin ten times
  num=sample(0:1,10,replace=TRUE)
  #Step 2: Total number of heads per student
  total=sum(num)
  #Step 3: Compute proportion
  prop=mean(num)
  #Step 4: Combine all students' results
  props=append(props,prop) 
}
mean(props)

[1] 0.5016

hist(props,prob=TRUE)
curve(dnorm(x,mean=mean(props)-0.05,sd=sqrt(var(props))),col="blue", add=TRUE)

What about this one?

We collected your data in the morning session. Let's explore it!

• Explore the collected data
• Compute the summary statistics
• Create simple plots
• Do simple linear regression

• gender : Male or Female
• shoe : Shoe size
• height : Height (in centimeters)

data=read.csv("http://pastebin.com/raw/MFYcN8Rk")

fix(data)

Typing the object name “data” will output the data to the R console.

The “fix” function also allows you to look at the data just like an Excel spreadsheet.

summary(data)

hist(data$shoe)

hist(data$shoe,col="red",xlab="Shoe Size (in cm)",
     main="Histogram of Shoe Size")

hist(data$height)

hist(data$height,col="green",xlab="Height (in cm)",
     main="Histogram of Height")

boxplot(data$height~data$gender,main="Boxplot",
        col=rainbow(2),xlab="Gender",ylab="Height")

boxplot(data$shoe~data$gender,main="Boxplot",
        col=rainbow(2),xlab="Gender",ylab="Shoe Size")

#Fit the regression model
lm=lm(data$height~data$shoe)

#plot the relationship
plot(x=data$shoe,y=data$height,xlab="Shoe Size",
     ylab="Height",main="Height vs. Shoe Size")

#add the regression line
abline(lm,col="red")

Our analysis shows that the profit of a Michael Bay movie increases by 3.21 million dollars for each added explosion.

The effect is very significant.

• Import and view data
• Summary statistics
• Visualizations (boxplot, pie chart, histogram)
• Linear regression

Shiny Application Example