• R is open source and freely available
• Works on all standard platforms (Mac, Windows, linux)
• Standard tool for many sciences (e.g., Computer Science, Physics)
• Easy to produce web-based slides using:
  • R-Studio
  • R-Markdown
• Traditional statistics tool with GUI front ends, as well as a flexible and powerful programming language
• Together with ggplot2 package can produce very nice data visualizations

The latest version of R as of this presentation is 3.2.3

1. Go to http://cran.r-project.org/mirrors.html
2. Pick a mirror site (e.g., http://watson.nci.nih.gov/cran_mirror/)
3. Select the Download for … link that matches your platform
   
4. Follow the download and install prompts

If you are using Windows, choose the 64-bit version unless you know you are running a 32-bit version of Windows

The latest version of R Studio as of this presentation is 0.99

1. Go to http://www.rstudio.com/products/RStudio/#Desk
2. Select Download RStudio Desktop, then pick your platform
3. Follow the download and install prompts
4. Watch their R Studio Overview video

• One can use the Console in R Studio to enter commands, functions, and data to perform statistical processes
• The Console is located in the bottom left of the R Studio screen
• When you see a grey box with code in it, you can copy and paste this into the Console and match the results to what the slides say under the grey box
• For example:

print("Statistics Rocks (roughly speaking)!")

## [1] "Statistics Rocks (roughly speaking)!"

• Later we’ll talk about using scripts to do more complicated things

1. Run R Studio
2. In the Console, type or copy the following:

install.packages(c("ggplot2","gcookbook","UsingR"))

If you are confused about how any function works in R, just type ?<functionname> for help – for example:

?install.packages

library(ggplot2)

## Registered S3 methods overwritten by 'ggplot2':
##   method         from 
##   [.quosures     rlang
##   c.quosures     rlang
##   print.quosures rlang

library(gcookbook)

ggplot(cabbage_exp, aes(x=Date, y=Weight, fill=Cultivar)) + 
  geom_bar(position="dodge",stat="identity")

• R comes with many standard distributions
  • Continuous: Exponential (exp), Normal (norm), Student \(t\) (t), Uniform (unif), Chi-Squared (chi), etc.
  • Discrete: Binomial (binom), Geometric (geom), Poisson (pois), etc.
• We can make use of a number of operations on distributions
  • r<dist>() to draw random numbers
  • q<dist>() to get quantile information
  • p<dist>() to get percentile information
  • d<dist>() to get density information

Seven random numbers from a binomial distribution, with \(n=100, p=0.25\)

rbinom(7,size=100,prob=0.25)

## [1] 28 33 32 20 28 27 31

Three ranom numbers \(~N(5,0.1)\)

rnorm(3,mean=5,sd=0.1)

## [1] 4.902377 4.940091 4.987001

qplot(rexp(100000,rate=2),binwidth=0.25)

• Let’s get Edgar Anderson’s dataset on Iris flowers
• This dataset is stored online as a comma separated file
• We can read such files directly using read.csv()
• We can aslo get a summary of all the variables in this dataset, as well as their quartile information (for numeric data) and level counts (for categorical data)
• With the correct package, there are commands to read many data formats
  • read.xlss() for Microsoft Excel
  • read.spss() for SPSS
  • read.xport() for SAS
  • Etc.

irisData = read.csv(
  "http://vincentarelbundock.github.io/Rdatasets/csv/datasets/iris.csv")
summary(irisData)

##        X           Sepal.Length    Sepal.Width     Petal.Length  
##  Min.   :  1.00   Min.   :4.300   Min.   :2.000   Min.   :1.000  
##  1st Qu.: 38.25   1st Qu.:5.100   1st Qu.:2.800   1st Qu.:1.600  
##  Median : 75.50   Median :5.800   Median :3.000   Median :4.350  
##  Mean   : 75.50   Mean   :5.843   Mean   :3.057   Mean   :3.758  
##  3rd Qu.:112.75   3rd Qu.:6.400   3rd Qu.:3.300   3rd Qu.:5.100  
##  Max.   :150.00   Max.   :7.900   Max.   :4.400   Max.   :6.900  
##   Petal.Width          Species  
##  Min.   :0.100   setosa    :50  
##  1st Qu.:0.300   versicolor:50  
##  Median :1.300   virginica :50  
##  Mean   :1.199                  
##  3rd Qu.:1.800                  
##  Max.   :2.500

• Use the <dataset-name>$<variable-name> to get the data from a specific column
• Compute basic statistics on columns, e.g, mean()

mean(irisData$Sepal.Length)

## [1] 5.843333

• Get all the levels of a categorical variable:

levels(irisData$Species)

## [1] "setosa"     "versicolor" "virginica"

A 95% confidence interval of a mean using the standard normal distribution:

\[ \bar{x} \pm z^\star \cdot \frac{s}{\sqrt{n}} \]

n = dim(irisData)[1]                 # Sample Size
xb = mean(irisData$Sepal.Length)     # x_bar
s  = sqrt(var(irisData$Sepal.Length)) # standard deviation
z  = qnorm(0.975)                    # Critical value for 95%
xb - z*s/sqrt(n)
xb +z*s/sqrt(n)

## [1] 5.975849

## [1] 5.710818

R comes with an easy way to get confidence intervals using the Student \(t\) distribution:

t.test(irisData$Sepal.Length)$conf.int

## [1] 5.709732 5.976934
## attr(,"conf.level")
## [1] 0.95

For a normal distribution, we load the UsingR package:

library(UsingR)
simple.z.test(irisData$Sepal.Length,
              sigma=sqrt(var(irisData$Sepal.Length)))

## [1] 5.975849 5.710818

• To get a sequence from 1 to 10 stepping by 1, use 1:10
• To make your own list of things, use c()
• To count the values in the list, use length()

1:10

##  [1]  1  2  3  4  5  6  7  8  9 10

c(3,-1.2,44,103.009)

## [1]   3.000  -1.200  44.000 103.009

length(c(3,-1.2,44,103.009))

## [1] 4

• To sum the values in a list of numbers, use sum()
• To produce a running, cumulative sum of a list, use cumsum()
• The same idea applies to product, prod() and cumprod()

sum(rnorm(10))   # Sum 10 numbers drawn from std. norm

## [1] -5.292483

cumsum(1:10)     # Running sum of numbers from 1:10

##  [1]  1  3  6 10 15 21 28 36 45 55

cumprod(c(2,4,10))

## [1]  2  8 80

• You can type commands into the console

• But for anything even moderately sophisticated, you’ll probably want to create a source file

• An R source file is a text-readable file with R commands in them that can be run any time

• There are many advantages, not the least of which is being able to give the file to someone else to run

• Also, you’ll need to turn in a source file for homework

• You can use any text editor, but we’ll use R Studio

• Go to File \(\rightarrow\) New File \(\rightarrow\) R Script

• Type commands into the file in the order that you would like them to be executed
  • Don’t forget to include any library() loads
  • The user running this file next may not have that library loaded
• Save the file when you are done
  • File \(\rightarrow\) Save As, then give the file a location and name that you like
  • Or just File \(\rightarrow\) Save if it already has a location and name
  • It is customary to give the file the extension .R or .r

To run the file in R Studio, either

• Press the Source button in the top right of the editor panel

• Or, in the console, type:

source('path/to/my/script.R')

It is always a good idea to make sure your script will run with your environment completely clean

1. From the top-right panel, select the Environment tab

2. Select the little broom icon to clean the environment, the confirm with Yes on the dialog

3. From the Session menu option, select Restart R and Clean Output

4. Source your file

• You will be given several R lab assignments throughout the semester

• Please submit these via git – there is a link in the assignment

• I will source your file and compare its output to what I expect to see

• I will also need any data files you used
  • If you refer to files on your local machine, I wont be able to read them

• numeric (numbers)
• character (strings)
• vectors (lists of one type)
• factors (categorical variables)
• lists (lists of arbitrary types)
• matrices (numeric 1D and 2D arrays)

• scalar numbers: integers and real values
• arithmatic operations on these

x = 3
y = 4.7
x/y + 2*x - 3

## [1] 3.638298

y > x

## [1] TRUE

• strings
• operations on strings

x = "hello"
nchar(x)

## [1] 5

gsub("he","HE-",x)

## [1] "HE-llo"

x = "hello"
y = "world"
paste(x,y)

## [1] "hello world"

paste(x,y,"turtle",sep=':')

## [1] "hello:world:turtle"

• What I called a “list” last week is really a “vector” in R
• R vectors aren’t just numeric
• However, they are the same type
• Creating a vector with a mixture of numeric and character values will force R to coerce the vector to character

c(1,2,9)

## [1] 1 2 9

c(1,2,"9")

## [1] "1" "2" "9"

x = c(-4,2,31)
3*x

## [1] -12   6  93

x[1]

## [1] -4

length(x)

## [1] 3

x = c(-4,2,3)
y = c(1,3,9)
z = 4*(1:3)
x*y

## [1] -4  6 27

x*y - z

## [1] -8 -2 15

x = c(-4,2,3)
names(x) <- c("Bob","Frank","Mindy")
x["Frank"]

## Frank 
##     2

x["Mindy"] = 0
print(x)

##   Bob Frank Mindy 
##    -4     2     0

x = 1:5
paste("String",x,sep='-')

## [1] "String-1" "String-2" "String-3" "String-4" "String-5"

x = seq(from=2,to=10,by=2)
print(x)

## [1]  2  4  6  8 10

x = c(x,-99)
x

## [1]   2   4   6   8  10 -99

x[7] = -98
x

## [1]   2   4   6   8  10 -99 -98

• A factor is like a vector, except for categorical values
• They can be created from vectors

x = c("good","good","bad","mediocre","good")
factor(x)

## [1] good     good     bad      mediocre good    
## Levels: bad good mediocre

factor(c(1,2,1,1,2,2,3,3,2,2))

##  [1] 1 2 1 1 2 2 3 3 2 2
## Levels: 1 2 3

x = list()
x[[1]] = 1
x[[2]] = "hello"
x[[3]] = c(-1,-2,-9)
x

## [[1]]
## [1] 1
## 
## [[2]]
## [1] "hello"
## 
## [[3]]
## [1] -1 -2 -9

x = list(3,1,4,5)
names(x) <- c("A","B","C","D")
x[["A"]]

## [1] 3

x$A

## [1] 3

A = 1:6
dim(A) <- c(2,3)
print(A)

##      [,1] [,2] [,3]
## [1,]    1    3    5
## [2,]    2    4    6

A[1,3]

## [1] 5

A[,2]

## [1] 3 4

A = 1:6
dim(A) <- c(2,3)
A[2,]; A[,3]

## [1] 2 4 6

## [1] 5 6

A[,2] = c(-99,-98)
print(A)

##      [,1] [,2] [,3]
## [1,]    1  -99    5
## [2,]    2  -98    6

A = 1:6
dim(A) <- c(2,3)
3*A

##      [,1] [,2] [,3]
## [1,]    3    9   15
## [2,]    6   12   18

A + A

##      [,1] [,2] [,3]
## [1,]    2    6   10
## [2,]    4    8   12

A = 1:6
dim(A) <- c(2,3)
A

##      [,1] [,2] [,3]
## [1,]    1    3    5
## [2,]    2    4    6

t(A)

##      [,1] [,2]
## [1,]    1    2
## [2,]    3    4
## [3,]    5    6

A = 1:6
dim(A) <- c(2,3)
A * A

##      [,1] [,2] [,3]
## [1,]    1    9   25
## [2,]    4   16   36

A %*% t(A)

##      [,1] [,2]
## [1,]   35   44
## [2,]   44   56

• Basic data type for statistical operations
• Implements a table
  • Variables stored in columns
  • Observations stored in rows
• Typically columns are named
• Values in each column must be the same type
  • factors
  • numeric vectors
  • character vectors
• Different columns may have different types

Student = c("Bob", "Sue", "Cat", "Lin")
NumberGrade = c(96, 82, 97, 74)
LetterGrade = factor(c("A","B","A","C"))
RosterData = data.frame(Student,NumberGrade,LetterGrade)
RosterData

##   Student NumberGrade LetterGrade
## 1     Bob          96           A
## 2     Sue          82           B
## 3     Cat          97           A
## 4     Lin          74           C

rd = data.frame( Student = c("Bob", "Sue", "Cat", "Lin"),
                 NumberGrade = c(96, 82, 97, 74),
                 LetterGrade = factor(c("A","B","A","C")) )
rd$NumberGrade

## [1] 96 82 97 74

rd$LetterGrade

## [1] A B A C
## Levels: A B C

# rd from last slide
dim(rd)

## [1] 4 3

rd[2,]

##   Student NumberGrade LetterGrade
## 2     Sue          82           B

# rd from two slides ago
summary(rd)

##  Student  NumberGrade    LetterGrade
##  Bob:1   Min.   :74.00   A:2        
##  Cat:1   1st Qu.:80.00   B:1        
##  Lin:1   Median :89.00   C:1        
##  Sue:1   Mean   :87.25              
##          3rd Qu.:96.25              
##          Max.   :97.00

• mode() describes how the data is stored
• class() describes what class interprets that data

c(mode(4), mode("hello"))

## [1] "numeric"   "character"

x = as.Date("2014-01-22"); mode(x)

## [1] "numeric"

class(x)

## [1] "Date"

• assign() allows you to assign values to a named variable
• get() allows you to get the value from a named variable
• rm() removes a named variable altogether

x = 2
assign("y",3)
print(c(get("x"),y))

## [1] 2 3

rm(x)
#print(x)

x = read.table('http://cs.ucf.edu/~wiegand/idc6700/datasets/color-cookbook-eg.txt', 
               header=T)
head(x) 

##   cond1 cond2 yval
## 1     A     I  2.0
## 2     A     J  2.5
## 3     A     K  1.6
## 4     A     L  0.8
## 5     B     I  2.2
## 6     B     J  2.4

We’ll be learning a lot more about R as the semester progresses, including a number of lab assignments. But there’s also a lot of on-line materials, including:

• The online Cookbook for R, http://www.cookbook-r.com
• A simple article in Computer World introducing R, http://www.computerworld.com/article/2497143/business-intelligence-beginner-s-guide-to-r-introduction.html
• Another introductory article in Computer Word about getting your data into R, http://www.computerworld.com/article/2497164/business-intelligence/beginner-s-guide-to-r-get-your-data-into-r.html
• The R-Bloggers site, http://www.r-bloggers.com
  • This includes a two-part introduction to R series, http://rtutorialseries.blogspot.com/2009/10/r-tutorial-series-introduction-to-r_11.html

R Cookbook, by Paul Teetor

R Graphics Cookbook, by Winston Chang