Introduction to the

R Project for Statistical Computing

Adjunct Associate Professor
School of Integrative Plant Sciences, Section of Soil & Crop Sciences
Cornell University.

D G Rossiter

November 27, 2021

Copyright © 2014-2016, 2019-2021 David G Rossiter

All rights reserved. Reproduction and dissemination of the work as a whole (not parts) freely permitted if this original
copyright notice is included. Sale or placement on a web site where payment must be made to access this document is strictly
prohibited. To adapt or translate please contact the author ([email protected]).
Introduction to R 1

Topics – Part 1

1. The R Project for Statistical Computing: what and why?

2. Installing R and RStudio

3. Interacting with R

4. The S language: expressions, assignment, functions

5. The R help system, R manuals, on-line R help

6. Finding, installing and loading contributed packages

7. Finding and loading example datasets

D G Rossiter
Introduction to R 2

The R Project for Statistical Computing: what and why?

• R is an open-source environment for statistical computing, data manipulation

and visualisation;

• Statisticians have implemented over 2 000 specialised statistical procedures as

contributed packages;

• R and its packages are freely-available over the internet;

• R runs on many operating systems, including Microsoft Windows, Unix © and

derivatives Mac OS X and Linux;

• R is fully programmable, with its own modern computer language, S;

• Repetitive procedures can be automated by user-written scripts, functions or

packages;

• ...

D G Rossiter
Introduction to R 3
• ...

• R is supported by comprehensive technical documentation, user-contributed tutorials

and textbooks; these all have freely-available R code

• R is the lingua franca ([U+666E][U+901A][U+8BDD]) of the computational statistics

world.

• R can import and export in MS-Excel, text, fixed and delineated formats (e.g. CSV),
with databases . . . ;

• R is a major part of the open source and reproducible research movement for
transparent and honest science.

D G Rossiter
Introduction to R 4

Installing R and RStudio

• R is the the computing environment; RStudio is an Integrated Development

Environment (IDE) which makes using R easier

• Install R first; it can run outside RStudio

– The Comprehensive R Archive Network (CRAN): http://cran.r-project.org/

to download R, packages and documentation
– link “Download R for . . . ” (Linux, Mac OS/X, Windows)
– Install the “base” version

• Install RStudio from its home page http://www.rstudio.com/

– link “Download RStudio” desktop open-source version

• Start RStudio; it will automatically start R.

D G Rossiter
Introduction to R 5

RStudio Features (1/2)

• R console

– enter R commands here, see text output

• Code editor

– write one or more R commands, pass the commands to the console and see the text
output there
– advantage: can edit and re-run
– can save the script to reproduce the analysis

• Graphics viewer (“Plots”)

– shows output of commands that produce figures

– can save for printing or inclusion in reports

D G Rossiter
Introduction to R 6

RStudio Features (2/2)

• Workspace viewer

– shows the objects in your workspace

• File manager

• History viewer

• Package manager

– install (from CRAN) and load (in your workspace) additional packages

• Integrated help system

• Project manager

– can switch between data analysis projects, each in its own directory

D G Rossiter
Introduction to R 7

RStudio Screenshot

D G Rossiter
Introduction to R 8

Basic interaction with the R console

• > is a prompt: R is waiting for input:

>

• You can type directly after the prompt; press the Enter to submit the command to R

• If a command is not syntactically-complete, R will show the continuation prompt:

• When the command is complete, R will execute

• Better: type a command in the code editor and click the Run button or press
Alt+Enter to pass the command to the console

• Text output (if any) will appear in the console; figures will appear the graphics window

D G Rossiter
Introduction to R 9

First interaction with the console

Draw 100 normally-distributed random numbers (µ = 0, σ 2 = 1), summarize them:
> summary(rnorm(100))

Min. 1st Qu. Median Mean 3rd Qu. Max.

-2.36000 -0.85350 -0.06113 -0.14620 0.58610 2.47700

Draw another set of 100 and graph

them as a histogram:
> hist(rnorm(100))

Your results will be different – why?

D G Rossiter
Introduction to R 10

The S language

1. Origin; R vs. S

2. Expresssions

3. Assignment and the workspace

4. Functions

D G Rossiter
Introduction to R 11

Origin of S

• The language implemented in the R environment is S

• Developed at Bell Laboratories (USA) in the 1980’s (John Chambers etc.)

• Designed for “programming with data”, including statistical analysis

• Line between “user” and “programmer” purposely blurred

• Syntax similar to ALGOL-like programming languages (C, Pascal, and Java . . . )

• Operators,functions and methods are generally vectorized; vector and matrix

operations are expressed naturally

• Statistical models specified with a standard notation

D G Rossiter
Introduction to R 12

Origin of R
• 1990–1994 Ross Ihaka, Robert Gentleman at Univ. Auckland (NZ), for own teaching
and research

• Syntax from S, internals from Scheme (a LISP-like functional programming language)

• 1997 Kurt Hornik and Fritz Leisch establishe the CRAN (Comprehensive R Action
Network) archive at TU Vienna

• 2000 V1.0 official release

• R Core Team of developers (Ripley, Dalgaard, Lumley, Tierney, Plummer . . . )

• S3 and then S4 object-oriented systems (V2, V3)

• Independent package developers

• 2015 Microsoft aquires Revolution Analytics

https://mran.revolutionanalytics.com – still open-source but “industrial-level”
support for Big Data projects
D G Rossiter
Introduction to R 13

Expressions
R can be used as a command-line calculator; these S expressions can then be used
anywhere in a statemnt.

> 2*pi/360

[1] 0.0174533

> 3 / 2^2 + 2 * pi

[1] 7.03319

> ((3 / 2)^2 + 2) * pi

[1] 13.3518

D G Rossiter
Introduction to R 14

Assignment
Results of expressions can be saved as objects in the workspace.

There are two (equivalent) assignment operators:

> rad.deg <- 2*pi/360

> rad.deg = 2*pi/360

By default nothing is printed; but all of these:

> (rad.deg <- 2*pi/360)

> rad.deg
> print(rad.deg)

give the same output:

[1] 0.0174533

D G Rossiter
Introduction to R 15

Workspace objects

• Create by assignment

• May be complex data structures (see ‘methods’)

– In the example below we use the c ‘catenate; build a chain’ function to build a vector

• List with ls ‘list’ or objects functions

• Delete with the rm (remove) function

> (heights <- c(12.2, 13.1, 11.9, 15.5, 10.9))

[1] 12.2 13.1 11.9 15.5 10.9

> ls()

[1] "heights"

> rm(heights); ls()

character(0)
D G Rossiter
Introduction to R 16

Functions and Methods

Most work in S is done with functions or methods:

1. Method or function name; any arguments between parentheses ( )

2. Argument list
(a) Required
(b) Optional, with defaults
(c) positional and/or named

These usually return some values, which can be complex data structures

D G Rossiter
Introduction to R 17

Example of a function call

Function name: rnorm (sample from a normal distribution)
Required argument: n: number of sampling units
Optional arguments: mean, sd

> rnorm(20)

[1] 0.388120 0.051022 -1.090701 0.155238 1.725087 2.011053 -2.122989 -0.685271

[9] -0.112195 0.876962 0.053067 -1.099789 0.299773 0.147167 -0.808183 -0.403877
[17] 1.173150 -1.557166 0.257684 -0.061434

> rnorm(20, mean=180)

[1] 180.99 180.89 180.64 181.64 179.45 179.90 179.04 179.62 178.94 180.66 179.35
[12] 180.16 179.31 179.66 178.05 180.07 181.58 179.37 179.08 180.21

> rnorm(20, mean=180, sd=10)

[1] 171.90 179.90 189.82 191.80 182.41 187.19 162.89 202.09 185.78 188.01 174.15
[12] 183.09 158.83 175.42 166.60 188.93 181.84 177.15 167.56 177.75

D G Rossiter
Introduction to R 18

The R help system, R manuals, on-line R help

1. R help

2. R manuals

3. on-line R help

D G Rossiter
Introduction to R 19

Help on functions or methods

Each function or method is documented with a help page, accessed by the help function:

> help(rnorm)

or, for short:

> ?rnorm

In R Studio can also search in the Help tab.

D G Rossiter
Introduction to R 20

Output from the help function

• Title and package where found

• Description

• Usage (how to call)

• Arguments (what each one means, defaults)

• Details of the algorithm

• Value returned

• Source of code

• References to the statistical or numerical methods

• See Also (related commands)

• Examples of use and output

D G Rossiter
Introduction to R 21

Example help page (1/2)

D G Rossiter
Introduction to R 22

Example help page (2/2)

D G Rossiter
Introduction to R 23

R manuals

• Included in the R distribution

• Access in R Studio with the Help tab or Help | R help menu item

• Six manuals; the first two are most relevant to an end-user

– An Introduction to R – somewhat difficult reading but packed with information

– R Data Import/Export
– R Installation and Administration
– The R Language Definition
– Writing R Extensions
– R Internals

• Reference cards (“cheatsheets”)

• FAQ

D G Rossiter
Introduction to R 24

on-line R help

• R task views

• StackOverflow R tags

• RSeek: http://www.rseek.org/

• User-written manuals, reference cards etc.: http://cran.r-project.org/, link

“Contributed”

D G Rossiter
Introduction to R 25

StackOverflow
URL: http://stackoverflow.com/questions/tagged/r: “Stack Overflow is a
question and answer site for professional and enthusiast programmers.”

Q&A tagged; the “R” tag is used for R questions.

For statistics questions, see http://stats.stackexchange.com: “Cross Validated is a

question and answer site for people interested in statistics, machine learning, data analysis,
data mining, and data visualization.”

You can post questions, always with small, reproducible examples – often writing those
examples will give you the solution yourself!

D G Rossiter
Introduction to R 26

StackOverflow R tags

D G Rossiter
Introduction to R 27

RSeek results

D G Rossiter
Introduction to R 28

R Task Views
Some applications are covered in Task Views, on-line at
http://cran.r-project.org/web/views/index.html.
These are a summary by a task maintainer of the facilities in R (e.g., which packages and
functions to use) to accomplish certain tasks.
Examples:

• Analysis of Spatial Data

http://cran.r-project.org/web/views/Spatial.html

• Handling and Analyzing Spatio-Temporal

Datahttps://cran.r-project.org/web/views/SpatioTemporal.html

• Multivariate Statistics
http://cran.r-project.org/web/views/Multivariate.html

• Analysis of Ecological and Environmental Data

http://cran.r-project.org/web/views/Environmetrics.html
D G Rossiter
Introduction to R 29

Contributed packages and example datasets

A major strength of R is the availability of user-contributed packages; 18 498 as of
27-Nov-2021!

You don’t need all of them!

These are often described in journal articles, books or technical reports, e.g.,

Baddeley, A., & Turner, R. (2004). spatstat: An R Package for Analyzing Spatial
Point Patterns. Journal of Statistical Software, 12(6). Retrieved from
http://www.jstatsoft.org/v12/i06

Fox, J. (2002). An R and S-PLUS Companion to Applied Regression. Newbury Park:

Sage. (the car package)

Diggle, P. J., & Ribeiro Jr., P. J. (2007). Model-based geostatistics. Springer. (the
geoR package)

D G Rossiter
Introduction to R 30

Installing packages

1. Find list at http://cran.r-project.org/; link “Packages”, link “Table of available

packages, sorted by name”

2. In RStudio: “Packages” pane, “Install” button; enter the names of the packages to install

3. Also check “Install dependecies” – most packages depend on others to also be on the
system

4. The first time you will be prompted to pick a repository, also known as mirror – R is
hosted at 100’s of sites around the world; they should all have the same packages

D G Rossiter
Introduction to R 31

Loading packages
The library and require functions (almost equivalent) load a package if it’s not already
in the workspace; they will also load dependencies (assuming these are installed on your
system):

> library(gstat) # `require(gstat)` is equivalent

Loading required package: gstat

Loading required package: sp

Using RStudio: “Packages” pane, check the packages to load.

D G Rossiter
Introduction to R 32

How do I find the method to do what I want?

• Mentioned in journal articles on subject of interest.

• Look at the help pages for methods you do know; they often list related methods.

• Search for keywords

– e.g., help.search("sequence") lists methods to generate sequences, vectors of

sequences, and sequences of dates for time-series analysis.

• Look at the Task Views http://cran.r-project.org/src/contrib/Views/

• Search the contributed documentation at CRAN

• Find a textbook that uses R

D G Rossiter
Introduction to R 33

Example data

• Base R includes a datasets package with many example datasets

• Most packages also include example data, which are used to explain the packages’
functions and methods

• When a package is loaded, so is its example data

• List datasets with data(); this is shown in a file frame

• Once you know the dataset name, see its documentation with ? or help

• To load into the workspace, use the data function with the dataset name

D G Rossiter
Introduction to R 34

Example
> data()
> ?CO2
> data(CO2)
> library(sp)
> data(package="sp")
> ?meuse
> data(meuse)
CO2 is a dataset in the
datasets package

meuse is a dataset in the

sp “classes and methods
for spatial data” package

D G Rossiter
Introduction to R 35

Topics – Part 2

1. Data types: logical, numeric (integer, double, complex), character, lists

2. Arrays, matrices and dataframes

3. Vectorized operations; applying functions over arrays

4. Matrix and dataframe manipulation

5. Logical expressions

6. Importing and exporting data

7. Summarizing data

8. Basic statistical functions

9. Specifying statistical models; the lm (linear models) function

D G Rossiter
Introduction to R 36

Data types and structures

• Data types: logical, numeric (integer, double, complex), character, lists

• Arrays, matrices and dataframes

• Vectorized operations; applying functions over arrays

• Matrix and dataframe manipulation

• Factors (categorical variables)

D G Rossiter
Introduction to R 37

Basic data types

• All objects in S have a data type

• Operators and functions understand these

• Some basic types: logical; integer; double; character; list; expression;

function

• logical; integer; double; character are all vectors with one or more elements

• lists can combine any objects

D G Rossiter
Introduction to R 38

Derived data types

These are basic types with some additional attributes appropriate to the derived type.

Examples:

• a array is a vector with a dim “dimensions” attribute

• a matrix is a 2-D array

• a dataframe is a matrix with column (field) names and row.names

D G Rossiter
Introduction to R 39

Vectorized operations
S works on vectors and matrices as with scalars, with natural extensions of operators,
functions and methods.

> (sample <- seq(1, 10) + rnorm(10))

[1] -0.1878978 1.6700122 2.2756831 4.1454326

[5] 5.8902614 7.1992164 9.1854318 7.5154372
[9] 8.7372579 8.7256403

The ten integers 1 ...10 returned by the call to the seq (sequence) method each have a
different random noise added to them; here the rnorm method also returns ten values.

If one of the vectors is shorter than the other, it is recycled as necessary:

> (samp <- seq(1, 10) + rnorm(5))

[1] -1.23919739 0.03765046 2.24047546 4.89287818

[5] 4.59977712 3.76080261 5.03765046 7.24047546
[9] 9.89287818 9.59977712

D G Rossiter
Introduction to R 40

Objects and classes

• S is an object-oriented computer language

• Everything in S is an object; every object has a class

• The class determines the way in which it may be manipulated

• Generic methods (e.g., summary, str) dispatch by the class

> class(seq(1:10)); class(seq(1,10, by=.01)); class(letters)

[1] "integer"
[1] "numeric"
[1] "character"

> class(diag(10)); class(iris); class(lm)

[1] "matrix"
[1] "data.frame"
[1] "function"
D G Rossiter
Introduction to R 41

Examples
> letters; letters + 3

[1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" "m" "n" "o" "p" "q" "r" "s"
[20] "t" "u" "v" "w" "x" "y" "z"

Error in letters + 3 : non-numeric argument to binary operator

> str(letters); str(diag(10)); str(iris)

chr [1:26] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j" "k" "l" ...

num [1:10, 1:10] 1 0 0 0 0 0 0 0 0 0 ...

'data.frame': 150 obs. of 5 variables:

$ Sepal.Length: num 5.1 4.9 4.7 4.6 5 5.4 4.6 5 4.4 4.9 ...
$ Sepal.Width : num 3.5 3 3.2 3.1 3.6 3.9 3.4 3.4 2.9 3.1 ...
$ Petal.Length: num 1.4 1.4 1.3 1.5 1.4 1.7 1.4 1.5 1.4 1.5 ...
$ Petal.Width : num 0.2 0.2 0.2 0.2 0.2 0.4 0.3 0.2 0.2 0.1 ...
$ Species : Factor w/ 3 levels "setosa","versicolor",..: 1 1 1 1 1 1 1 1 1 1 ...

D G Rossiter
Introduction to R 42

Matrices
> (cm <- c(35,14,11,1,4,11,3,0,12,9,38,4,2,5,12,2))

[1] 35 14 11 1 4 11 3 0 12 9 38 4 2 5 12 2

> dim(cm)

NULL

Initially, the vector has no dimensions; these are added with the dim function:

> dim(cm) <- c(4, 4)

> print(cm)

[,1] [,2] [,3] [,4]

[1,] 35 4 12 2
[2,] 14 11 9 5
[3,] 11 3 38 12
[4,] 1 0 4 2

> dim(cm)

[1] 4 4
D G Rossiter
Introduction to R 43

Matrix arithmetic
Many S operators can work directly on matrices; there are also typical matrix functions:

• +, -, *, / etc. work element-wise

• matrix multiplication: %*%

• transposition: t function

• inversion: solve function

• spectral decomposition: eigen function

• Singular Value Decomposition: svd function

• ...

D G Rossiter
Introduction to R 44

Data frames
The fundamental structure for statistical analysis; a matrix with:

1. named columns (roughly, database “fields”) and

2. (optionally) named rows (roughly, database “cases”):

We illustrate with one of R’s example datasets, provided in the base datasets package:

We first display the help file, then load the data, then view the data structure (field names
and types):

> ?trees
> data(trees)
> str(trees)

`data.frame': 31 obs. of 3 variables:

$ Girth : num 8.3 8.6 8.8 10.5 10.7 10.8 11 ...
$ Height: num 70 65 63 72 81 83 66 75 80 75 ...
$ Volume: num 10.3 10.3 10.2 16.4 18.8 19.7 ...

D G Rossiter
Introduction to R 45

Accessing fields in a data frame

Using the $ operator:

> summary(trees$Volume)

Min. 1st Qu. Median Mean 3rd Qu. Max.

10.2 19.4 24.2 30.2 37.3 77.0

This $ operator exposes the field name.

D G Rossiter
Introduction to R 46

Accessing a dataframe with matrix operators

The dataframe is just a special matrix, so:

> trees[1,] # one case, i.e. the first tree

Girth Height Volume

1 8.3 70 10.3

> trees[,2] # all cases (trees), second field (heights)

[1] 70 65 63 72 81 83 66 75 80 75 79 76 76 69
[15] 75 74 85 86 71 64 78 80 74 72 77 81 82 80
[ 29] 80 80 87

> trees[1,2] # one field of one case: height of first tree

[1] 70

D G Rossiter
Introduction to R 47
> trees[1:3,] # first three cases (trees), all fields

Girth Height Volume

1 8.3 70 10.3
2 8.6 65 10.3
3 8.8 63 10.2

> head(trees[,c(1,3)]) # first and third fields; `head' shows first six

Girth Volume
1 8.3 10.3
2 8.6 10.3
3 8.8 10.2
4 10.5 16.4
5 10.7 18.8
6 10.8 19.7

> trees[1,"Height"] # named field (i.e., matrix column)

[1] 70

D G Rossiter
Introduction to R 48

Factors

• Variables with a limited number of discrete values (categories) are called S factors.

• Internally they are stored as integers but each has a text name.

• They are handled properly by R functions and methods (they are not integers!).

– Unordered factors: no intrinsic order

– Ordered factors: intrinsic order relation, > etc. make sense

D G Rossiter
Introduction to R 49

Example of factors (1/2)

Suppose we have given three tests to each of three students, each with a numeric ID, and
we want to compare the students. We might enter the data frame as follows:

> student <- rep(c(700123, 131444, 201113), 3)

> score <- c(9, 6.5, 8, 8, 7.5, 6, 9.5, 8, 7)
> tests <- data.frame(cbind(student, score))
> str(tests)

'data.frame': 9 obs. of 2 variables:

$ student: num 700123 131444 201113 700123 131444 ...
$ score : num 9 6.5 8 8 7.5 6 9.5 8 7

The data type of student is numeric – this can lead to absurdities:

> lm(score ~ student, data=tests)

Coefficients:
(Intercept) student
6.682e+00 3.022e-06

Meaningless!
D G Rossiter
Introduction to R 50

Example (2/2)
Convert to a factor: the student number is just an ID; use as.factor:

> tests$student <- as.factor(tests$student)

> levels(tests$student)

[1] "131444" "201113" "700123"

> str(tests)

'data.frame': 9 obs. of 2 variables:

$ student: Factor w/ 3 levels "131444","201113",..: 3 1 2 3 1 2 3 1 2
$ score : num 9 6.5 8 8 7.5 6 9.5 8 7

> lm(score ~ student, data=tests)

(Intercept) student201113 student700123

7.3333 -0.3333 1.5000

This is a meaningful one-way linear model, showing the difference in mean scores of
students 201113 and 700123 from student 131444 (the intercept).

D G Rossiter
Introduction to R 51

Data manipulation
One of the strengths of R is the ability to manipulate data.

This is especially useful for automatic identification of suspected errors, outlier detection,
data transformations, subsetting on a factor . . .

D G Rossiter
Introduction to R 52

Subsetting on a logical expression

Find the tallest trees using the subset function:

> sort(trees$Height)
> subset(trees, Height >= 80)

[1] 63 64 65 66 69 70 71 72 72 74 74 75 75 75 76 76 77 78 79 80 80 80 80 80 81 81
[27] 82 83 85 86 87

Girth Height Volume

5 10.7 81 18.8
6 10.8 83 19.7
9 11.1 80 22.6
17 12.9 85 33.8
18 13.3 86 27.4
22 14.2 80 31.7
26 17.3 81 55.4
27 17.5 82 55.7
28 17.9 80 58.3
29 18.0 80 51.5
30 18.0 80 51.0
31 20.6 87 77.0

D G Rossiter
Introduction to R 53

Another way . . .
Can use logical expression as subscripts:

> (trees.tall <- trees[trees$Height >= 80 , ])

Girth Height Volume

5 10.7 81 18.8
6 10.8 83 19.7
9 11.1 80 22.6
17 12.9 85 33.8
18 13.3 86 27.4
22 14.2 80 31.7
26 17.3 81 55.4
27 17.5 82 55.7
28 17.9 80 58.3
29 18.0 80 51.5
30 18.0 80 51.0
31 20.6 87 77.0

D G Rossiter
Introduction to R 54

Identifying with a logical expression

Which are the tallest trees? Save indices for later use.

Use the which function to find the indices:

> (trees.tall.ix <- which(trees$Height >= 80))

> trees[trees.tall.ix, ]

[1] 5 6 9 17 18 22 26 27 28 29 30 31

Girth Height Volume

5 10.7 81 18.8
6 10.8 83 19.7
9 11.1 80 22.6
17 12.9 85 33.8
18 13.3 86 27.4
22 14.2 80 31.7
26 17.3 81 55.4
27 17.5 82 55.7
28 17.9 80 58.3
29 18.0 80 51.5
30 18.0 80 51.0
31 20.6 87 77.0
D G Rossiter
Introduction to R 55

More complicated logical expression

Find very thin trees:

> trees$hg <- trees$Height/trees$Girth

> sort(trees$hg)
[1] 4.223301 4.444444 4.444444 4.469274 4.500000 4.637681 4.682081 4.685714
[9] 4.723926 5.103448 5.182482 5.571429 5.633803 5.736434 5.897436 6.000000
[17] 6.250000 6.466165 6.589147 6.666667 6.666667 6.696429 6.818182 6.857143
[25] 6.991150 7.159091 7.207207 7.558140 7.570093 7.685185 8.433735
> summary(trees$hg)
Min. 1st Qu. Median Mean 3rd Qu. Max.
4.223 4.705 6.000 5.986 6.838 8.434
> sd(trees$hg)
> (trees.thin <- subset(trees, hg > (mean(trees$hg) + sd(trees$hg))))

Girth Height Volume hg

1 8.3 70 10.3 8.433735
2 8.6 65 10.3 7.558140
5 10.7 81 18.8 7.570093
6 10.8 83 19.7 7.685185
9 11.1 80 22.6 7.207207

D G Rossiter
Introduction to R 56

Visualizing the thin trees

> plot(trees$Height ~ trees$Girth, xlab="Girth", ylab="Height",
main="Thin trees", pch=20, cex=2,
col=ifelse(trees$hg > (mean(trees$hg) + sd(trees$hg)), "red", "darkgreen"))

Thin trees

●
●

85
●

●
●
● ●
80 ● ● ●●
●
●
●
●
Height

75

●● ●
● ●

● ●
●
70

●
●

●
65

●
●
●

8 10 12 14 16 18 20

Girth

D G Rossiter
Introduction to R 57

Import/Export
Reference: “R Data Import/Export”, R manual installed with R; available under Help menu

Most common interchange format for flat-files: Comma-Separated Values (CSV)

D G Rossiter
Introduction to R 58

CSV file import

Look at file in plain-text editor; note header line, and the , field separator
x,y,cadmium,elev,dist,om,ffreq,soil,lime,landuse
181072,333611,11.7,7.909,0.00135803,13.6,1,1,1,Ah
181025,333558,8.6,6.983,0.0122243,14,1,1,1,Ah
181165,333537,6.5,7.8,0.103029,13,1,1,1,Ah
181298,333484,2.6,7.655,0.190094,8,1,2,0,Ga
181307,333330,2.8,7.48,0.27709,8.7,1,2,0,Ah
181390,333260,3,7.791,0.364067,7.8,1,2,0,Ga

Import with read.csv:

> ds <- read.csv("test.csv")
> str(ds)

(results on next slide)

D G Rossiter
Introduction to R 59

Result of CSV file input

'data.frame': 6 obs. of 10 variables:
$ x : int 181072 181025 181165 181298 181307 181390
$ y : int 333611 333558 333537 333484 333330 333260
$ cadmium: num 11.7 8.6 6.5 2.6 2.8 3
$ elev : num 7.91 6.98 7.8 7.66 7.48 ...
$ dist : num 0.00136 0.01222 0.10303 0.19009 0.27709 ...
$ om : num 13.6 14 13 8 8.7 7.8
$ ffreq : int 1 1 1 1 1 1
$ soil : int 1 1 1 2 2 2
$ lime : int 1 1 1 0 0 0
$ landuse: Factor w/ 3 levels "Ah","Ga","Ga ": 1 1 1 2 1 3

Note that read.csv could determine field landuse is a factor.

But it was not able to do so for the factors ffreq, soil, lime. So, we have to convert:

> ds$ffreq <- as.factor(ds$ffreq)

D G Rossiter
Introduction to R 60

General file import

Very flexible read.table function:

• field delimeters

• integer / decimal separator

• Header line(s)

• Skip lines

• Specify data types

D G Rossiter
Introduction to R 61

File export
Very flexible write.table function.

> write.table(round(as.data.frame(kxy), 4), file="KrigeResult.csv",

sep=",", quote=T, row.names=F,
col.names=c("E", "N", "LPb", "LPb.var"))

There are also ways to export to spreadsheets, databases, images, GIS coverages . . .

D G Rossiter
Introduction to R 62

Statistical models in S

• Specified in symbolic form with model formulae

• These formulae are arguments to many statistical methods:

– lm (linear models)
– glm (generalised linear models)
– gstat methods such as variogram and krige

• Can also be used in other contexts:

– Base graphics methods such as plot and boxplot

– Trellis graphics methods such as levelplot

D G Rossiter
Introduction to R 63

Form of statistical models

• Left-hand side: (mathematically) dependent variable

• Formula operator ~

• Right-hand side: (mathematically) independent variable(s)

The simplest use is in simple linear regression:

> model <- lm(Volume ~ Height, data=trees); summary(model)

> # equivalent to: model <- lm(trees$Volume ~ trees$Height)

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -87.124 29.273 -2.98 0.00583
Height 1.543 0.384 4.02 0.00038

So, the tree volume is modelled as a linear function of the tree height.

D G Rossiter
Introduction to R 64

Model formula operators

• Additive effects: +

> model <- lm(Volume ~ Height + Girth, data=trees); summary(model)

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -57.988 8.638 -6.71 2.7e-07
Height 0.339 0.130 2.61 0.014
Girth 4.708 0.264 17.82 < 2e-16

• Interactions: *
> model <- lm(Volume ~ Height * Girth, data=trees); summary(model)

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 69.3963 23.8358 2.91 0.00713
Height -1.2971 0.3098 -4.19 0.00027
Girth -5.8558 1.9213 -3.05 0.00511
Height:Girth 0.1347 0.0244 5.52 7.5e-06
D G Rossiter
Introduction to R 65

• Crossing factors to a specified degree

> model <- lm(Volume ~ (Height + Girth)^2, data=trees)

In this case it’s the same as Height * Girth, because there are only two factors.

• Nested models: /

> model <- lm(Volume ~ Height / Girth, data=trees)

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -0.23114 7.74157 -0.03 0.9764
Height -0.41218 0.12316 -3.35 0.0023
Height:Girth 0.06070 0.00266 22.79 <2e-16

• Remove terms: -; for example, the intercept:

> model <- lm(Volume ~ Height -1, data=trees); summary(model)

Coefficients:
Estimate Std. Error t value Pr(>|t|)
Height 0.4047 0.0354 11.4 1.9e-12
D G Rossiter
Introduction to R 66

• Arithmetic in models: I() method if ambiguous

> model <- lm(Volume ~ I(Height^2), data=trees); summary(model)
> # otherwise would cross height with itself in the model

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -30.19193 15.02843 -2.01 0.05393
I(Height^2) 0.01038 0.00255 4.07 0.00033

D G Rossiter
Introduction to R 67

Updating models
Use the update function, previous LHS and RHS represented by .

> model <- lm(Volume ~ Height + Girth, data=trees); summary(model)

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -57.988 8.638 -6.71 2.7e-07
Height 0.339 0.130 2.61 0.014
Girth 4.708 0.264 17.82 < 2e-16

> model.2 <- update(model, . ~ . - Girth); summary(model.2)

Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -87.124 29.273 -2.98 0.00583
Height 1.543 0.384 4.02 0.00038

D G Rossiter
Introduction to R 68

Model objects: structure and access

• Modelling functions like lm return an object with a class

• You can look directly at the structure . . .

• . . . but it is preferable to use access methods such as coefficients, residuals,

fitted.

> model <- lm(Volume ~ log(Height), data=trees); class(model); str(model)

[1] "lm"

List of 12
$ coefficients : Named num [1:2] -461 114
..- attr(*, "names")= chr [1:2] "(Intercept)" "log(Height)"
$ residuals : Named num [1:31] -10.928 -2.511 0.939 -8.028 -19.005 ...
..- attr(*, "names")= chr [1:31] "1" "2" "3" "4" ...
...

D G Rossiter
Introduction to R 69

Using access functions

These extract information from the fitted model.

> summary(residuals(model))

Min. 1st Qu. Median Mean 3rd Qu. Max.

-2.09e+01 -9.77e+00 -2.51e+00 -4.73e-16 1.22e+01 3.11e+01

Other important access functions: summary, fitted, coef, anova, effects, vcov.

D G Rossiter
Introduction to R 70

Factors

• For categorical variables (can take only a defined set of values)

– unordered (nominal), e.g. land cover class

– ordered (ordinal), e.g. vegetation density class

• S calls these factors

• Methods (especially modelling) take appropriate action

• These are converted to contrasts in the design matrix of linear (and other) models

D G Rossiter
Introduction to R 71

Topics – Part 3

1. R base graphics

2. Scripts

3. User-defined functions

4. Programming in R: control structures

5. The R class structure; object-oriented programming

6. The ggplot2 graphics system

7. Some advanced statistical functions

8. Going further in R: task views, textbooks, tutorials

D G Rossiter
Introduction to R 72

R Graphics
R has a very rich visualization environment. There are (at least) four graphics systems:

1. Base graphics system: default graphics package (always loaded)

2. Trellis graphics: lattice package

3. “Grammar of Graphics” ggplot2 package

4. Grid graphics

R graphics are highly customizable; it is usual to write small scripts to get the exact
output you want.

Graphs may be displayed on screen or written directly to files for inclusion in documents.

D G Rossiter
Introduction to R 73

Base graphics

• Simple to learn

• Can make simple plots very easily

• Can also customize at will

• Some methods start a new plot, e.g. plot, hist, boxplot

• Other add to an existing (open) plot, e.g. points, lines, rug

D G Rossiter
Introduction to R 74

Example of default base graphics

> data(iris); with(iris, plot(Petal.Width ~ Petal.Length))

2.5
● ●●
● ●
●●●● ● ● ● ●
● ● ●
●●●● ● ●

2.0
●●●● ● ●
●● ● ●
●●
●● ● ●● ● ● ●
● ●
● ● ● ●
1.5
Petal.Width

● ●●● ●●●
● ● ●●● ●
● ●●●●
●●●●●
●● ● ● ●
● ●●
1.0

● ● ● ●●

●
0.5

●
● ●●● ●
●●● ●
● ●●
●●●●
●●
●●
●● ●
● ●●

1 2 3 4 5 6 7

Petal.Length

D G Rossiter
Introduction to R 75

Example of customized base graphics

Anderson Iris data

2.5
● ●●

Three species of Iris: ● ●

●
setosa ●●●● ● ● ● ●

●
versicolor ● ● ●

●
virginica ●●●● ● ●

2.0
●●●● ● ●

●● ● ●

●● ● ●● ● ● ●

● ●

● ● ● ●
Petal width (cm)

1.5

● ●●● ●●●

● ● ●●● ●

● ●●●●●●●

●● ● ● ●

● ●●
1.0

● ● ● ●●

●
0.5

● ●●● ●

●●● ●

● ●●●●●● ●

● ●●

1 2 3 4 5 6 7

Petal length (cm)

Centroids: mean (red) and median (blue)

D G Rossiter
Introduction to R 76

Code for previous graph

> attach(iris)
> plot(Petal.Length, Petal.Width, pch=20, cex=1.2,
xlab="Petal length (cm)", ylab="Petal width (cm)",
main="Anderson Iris data",
col=c("slateblue", "firebrick", "darkolivegreen")[as.numeric(Species)]
)
> abline(v=mean(Petal.Length), lty=2, col="red")
> abline(h=mean(Petal.Width), lty=2, col="red")
> abline(v=median(Petal.Length), lty=2, col="blue")
> abline(h=median(Petal.Width), lty=2, col="blue")
> grid()
> points(mean(Petal.Length), mean(Petal.Width), cex=2, pch=23, col="black", bg="red")
> points(median(Petal.Length), median(Petal.Width), cex=2, pch=23,
col="black", bg="blue")
> title(sub="Centroids: mean (green) and median (gray)")
> text(1, 2.4, "Three species of Iris", pos=4, col="navyblue")
> legend(1, 2.4, levels(Species), pch=20, bty="n",
col=c("slateblue","firebrick", "darkolivegreen"))
> detach(iris)

Note that plot starts a new graph; all the others add elements to the plot.
D G Rossiter
Introduction to R 77

Trellis graphics
An R implementation of the trellis graphics system developed at Bell Labs by Cleveland is
provided by packakge lattice.

It is especially intended for multivariate visualization

• Harder to learn than R base graphics

• Can produce higher-quality graphics, especially for multivariate visualisation when the
relationship between variables changes with some grouping factor; this is called
conditioning the graph on the factor

• It uses model formulae similar to the statistical formulae to specify the variables to be
plotted and their relation in the plot.

• Multiple items on one plot are specified with user-written panel functions

D G Rossiter
Introduction to R 78

Example of trellis graphics

All species Split by species
setosa ●

versicolor ● 1 2 3 4 5 6 7
virginica ●
setosa versicolor virginica

2.5 ● ●●

2.5 ● ●● ● ●

● ● ●●●● ●●● ●

●●●● ● ● ● ● ●● ●

● ● ● ●●●●● ●

●●●● ● ●
2.0 ●●●● ●●

2.0 ●●●● ● ● ●●● ●

●● ● ● ● ●●● ●●●● ●
●● ● ●● ● ● ●
● ●
● ●
●● ● ●

Petal.Width
● ● ● ●
1.5 ● ●●●● ●●
Petal.Width

1.5 ● ●●● ●●●

● ●●●● ●
● ● ●●● ●
● ●●●●●●●
● ●●●●●●●
●●●● ●
●● ● ● ●
● ●●
● ●●
1.0 ●●● ●●
1.0 ● ● ● ●●

●
●

0.5 ● 0.5 ●

● ●●● ● ●●●●●

●●● ● ●●●●

● ●●●●●● ● ●●●
●●
●●
●●●●●

● ●● ● ●●

0.0 0.0

1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7
Petal.Length Petal.Length

Note the right plot: it has been conditioned on a factor, namely the species.

D G Rossiter
Introduction to R 79

Code for previous graph

> xyplot(Petal.Width ~ Petal.Length, data=iris, groups=Species, auto.key=T)
> xyplot(Petal.Width ~ Petal.Length | Species, data=iris, groups=Species)

Note the | in the formula; this means “conditioned on”.

D G Rossiter
Introduction to R 80

Grammar of graphics

• A completely new way to think about composing statistical graphs

• Text: Wickham, H., 2009. ggplot2: Elegant Graphics for Data Analysis, Use R! Springer.

• Web site: http://ggplot2.org

• How are statistical graphics a “grammar” ([U+8BED][U+6CD5])?

– a mapping from data to aesthetic attributes (colour, shape, size) . . .

– . . . of geometric objects (points, lines, bars).
– data may also be statistically-transformed
– the graph must be drawn on a coordinate system
– subsets of the data can be shown in sub-windows (“faceting”)

• R code to specify can be quite complex

• But the qplot “quick plot” method can be used for many simple cases (analogous to
plot of base graphics).
D G Rossiter
Introduction to R 81

Example of ggplot2 graphics

● ●

● ●
●
● ●●
●●
●
● ●● ●
●● ●

● ●
● ●
2.5 ●
●●
●
2.5 ●
●●
● ●●
●● ●
●
● ●
●●● ● ●● ●
● ●●
● ● ● ●
●● ● ●
●
● ●● ● ● ●● ●
● ●
● ● ●
● ● ● ● ● ●
● ●● ● ● ●
●
● ● ● ● ●
● ● ● ● ● ●
● ● ● ●
● ● ●● ●
● ● ● ● ● ●
● ● ●● ● ffreq
log10(lead)

log10(lead)
● ●
● ●●
● ●
●● ●
● ● ● ● ●●
● ●● ●
● ●● ● ● ● ● ● ● 1
●● ● ●
● ●● ●●
● ● ●● ● ●
●
● ● ●
● 2
●● ● ● 2.0 ● ● ● ● ● ●
● ● ●
● ● ● ● ●● ●
● 3
● ●
2.0 ● ● ●
● ● ● ●● ● ● ●
● ● ● ● ● ● ● ● ●
● ●
●
● ● ●
● ● ● ●●
● ● ● ●●● ● ●
● ● ● ●● ● ● ● ● ● ● ●
● ● ●
● ● ● ● ●
●● ● ● ● ● ●
● ● ●
●●● ● ● ● ● ●
● ● ● ● ● ●● ● ●● ● ● ●
● ● ● ● ● ●
●
● ● ●
● ● ●
● ● ● ●
●● ● ●● ● ● ● ●
● ● ● ●
●
● ●
1.5
●
●

1.5

0 250 500 750 1000 0 250 500 750 1000

dist.m dist.m

D G Rossiter
Introduction to R 82

Code for previous graph

> library(sp); data(meuse)
> qplot(x = dist.m, y = log10(lead), data = meuse,
+ geom = c("point", "smooth", method='loess')
> qplot(x = dist.m, y = log10(lead), data = meuse,
+ colour = ffreq, geom = c("point", "smooth"), method="loess")

• Data is the meuse dataframe

• two geometries are specified: (1) the points (a scatterplot); (2) a smooth line

• the coordinate system is by default a scatterplot (x-y plot)

• the x and y axes are the two named variables; the Pb content is log-transformed

• in the right-hand graph the points are coloured by a categorical variable (flood
frequency class)

• the smooth line and confidence limits are computed by locally-adjusted least squares
D G Rossiter
Introduction to R 83

Grid graphics
A low-level graphics programming language by Paul Murrel. lattice is written in grid.
Allows fine control of graphic output.

Complete information on author’s R graphics page:

http://www.stat.auckland.ac.nz/~paul/grid/grid.html

and in his book:

http://www.stat.auckland.ac.nz/~paul/RGraphics/rgraphics.html

D G Rossiter
Introduction to R 84

Programming R
R is a full-featured, modern programming language. This can be accessed four ways, in
increasing level of complexity:

S was developed by Chambers for “programming with data”

1. Commands: at the > prompt, typed or cut-and-paste

• These can use control structures for looping, conditional execution, and repetition

2. User-written scripts

3. User-defined functions

4. User-contributed packages

D G Rossiter
Introduction to R 85

Control structures
S has ALGOL-like control structures:

• if ...else

• for; note that vectorized functions or methods often are preferable

• while, repeat

• break, next

and within an expression:

• the ifelse function

D G Rossiter
Introduction to R 86

Example of the ifelse function

Here it is used to select a plotting colour:

> x <- rnorm(100); y <- runif(100, -3, 3)

> plot(y ~ x, asp=1, col=ifelse(y > x, "red", "green"), pch=20, cex=1.5)
> abline(0, 1, lty=2)

3
● ● ● ●●
● ● ●● ●
● ●●
● ●●
● ● ●
2
●
●● ●
● ●
● ● ● ●
● ● ● ● ●
● ● ●
1

● ● ●
● ● ● ●
● ●●
● ● ●●●
● ●
●
0
y

●● ●
●● ● ● ●
●
●
●●
● ●● ● ● ●
−1

●
●
● ● ● ●
●
● ●●
−2

●
●
● ●●
● ● ●
●
●● ● ●
●
−3

● ● ●

−3 −2 −1 0 1 2 3

D G Rossiter
Introduction to R 87

Example of the while control structure

For some simulation we want to draw a sample from the normal distribution but make sure
there is an extreme value, so we repeat the sampling until we get what we want:

> while (max(abs(sample <- rnorm(100))) < 3) print("No extreme")

> range(sample)

[1] "No extreme"

[1] "No extreme"
[1] "No extreme"

[1] -3.2648 2.5457

D G Rossiter
Introduction to R 88

Why use scripts?

• For reproducible processing

– Especially for complicated graphics

– Also for multi-step analyses
– For simulations where each run is different, due to randomness

• Can document the steps internally (as S comments)

D G Rossiter
Introduction to R 89

Writing and running scripts

1. Prepare script in some editor

• Editor component of IDE, e.g., RStudio
• Plain-text editor (no formatting!)
• Editor built into R: some help with syntax, commands
• Emacs + ESS (“Emacs speaks statistics”) http://ess.r-project.org/

2. Run with the source function or via editor commands

• e.g., RStudio “Run Lines” or “Run Region” menu commands

D G Rossiter
Introduction to R 90

RStudio screenshot with script and console

D G Rossiter
Introduction to R 91

Example
1. Enter the following in a plain text file:
# draw two independent normally-distributed samples
x <- rnorm(100, 180, 20); y <- rnorm(100, 180, 20)
# scatterplot
plot(x, y)
# correlation: should be 0
cor.test(x, y, conf=0.9)

2. Save with name e.g. test.R (convention: .R extension)

3. In R, source the file (or send from the editor):

> source("test.R")

t = -0.1925, df = 98, p-value = 0.8477

alternative hypothesis: true correlation is not equal to 0
90 percent confidence interval:
-0.18433 0.14650
sample estimates:
cor
-0.019446
D G Rossiter
Introduction to R 92

A more complicated example

Enter this in a script file; save as test.R.

# see how correlation coefficients are distributed in uncorrelated random samples

m <- 1000 # number of runs
n <- 100 # size of random samples
results <- rep(0, m)
for (i in 1:m) {
x <- rnorm(100); y <- rnorm(100) # default mu=0, sigma=1
results[i] <- cor(x, y)
}
summary(results)
tmp <- qplot(results, binwidth=0.02)
print(tmp + geom_bar(colour="white", fill="darkgreen", binwidth=0.02) + geom_rug())

Run the script:

> source('test.R')

The script can be run several times, also with different numbers of runs and sample sizes,
to compare the results.

D G Rossiter
Introduction to R 93

Results
Min. 1st Qu. Median Mean 3rd Qu. Max.
-0.382500 -0.071820 -0.001889 -0.001115 0.072200 0.355600

D G Rossiter
Introduction to R 94

User-defined functions

• These are like R built-in functions but simpler

• Defined as objects in the workspace (not in the system)

• Why?

– R may not have a function or method to compute what you want

– You want to expand a script with arguments to apply the script to any suitable
object

D G Rossiter
Introduction to R 95

Simple example of user-defined function

There is no R function to compute the harmonic (geometric) mean of a vector, but we
can define it easily enough. For a vector v with n elements:

 1/n
Y
v̄h =  vi
i=1...n

This is computed by taking logarithms, dividing by the length, and exponentiating.

The function function is used to define a function (!); it can then be assigned to an
object in the workspace. The function has one argument, here named v:

> hm <- function(v) exp(sum(log(v))/length(v))

> class(hm)
> hm(1:99); mean(1:99)

[1] "function"
[1] 37.6231
[1] 50

D G Rossiter
Introduction to R 96

A better version
A function should check for valid inputs. This shows the use of the if, else if, else
control structure:

> hm <- function(v) {

if (!is.numeric(v)) {
print("Argument must be numeric"); return(NULL)
}
else if (any(v <= 0)) {
print("All elements must be positive"); return(NULL)
}
else return(exp(sum(log(v))/length(v)))
}
> hm(letters)
> hm(c(-1, -2, 1, 2))
> hm(1:99)

[1] "Argument must be numeric"

NULL
[1] "All elements must be positive"
NULL
[1] 37.6231

D G Rossiter
Introduction to R 97

Another example
The “correlation of two random normal vectors” script can be converted to a function; the
arguments are the number of runs and sample size:

> corr.two.random.normal <- function(m =1000, n=100) {

+ results <- rep(0, m)
+ for (i in 1:m) {
+ x <- rnorm(100); y <- rnorm(100) # default mu=0, sigma=1
+ results[i] <- cor(x, y)
+ }
+ summary(results)
+ tmp <- qplot(results, binwidth=0.02)
+ print(tmp + geom_bar(colour="white", fill="darkgreen", binwidth=0.02) + geom_rug())
+ }

The function is now defined in the workspace; to call it:

> corr.two.random.normal() # with defaults

> corr.two.random.normal(256, 20) # specify m and n

Try it! The second histogram will be much more erratic than the first.

D G Rossiter
Introduction to R 98

Some advanced statistical functions

This is a very small sample of what is available.

D G Rossiter
Introduction to R 99

Modelling

• Non-linear model fitting: nls

• Non-linear mixed-effects models: nlme package

• Generalized linear models (GLM): glm

• Robust fitting of linear models: lqs, lm.ridge etc.

• Local (smooth) fitting: loess

• Stepwise regression: step

• Regression trees: trees, rpart packages

• Principal component, partial least squares: prcomp, pls packages

• Random forests: randomForest package

D G Rossiter
Introduction to R 100

Simulation

• Bootstrapping: boot package

D G Rossiter
Introduction to R 101

Time and space

• Time-series analysis: ts, arima etc.

• Spatially-explicit objects: sp package

• Geostatistics: gstat, geoR, spatial packages

• Space-time geostatistics: spacetime package

• Point-pattern analysis: spatstat, spatial packages

• Areal spatial data analysis (like GEODA): spdep package

• Interface to GIS: rgdal, RSAGA packages

• Image processing: raster package

D G Rossiter
Introduction to R 102

Resources for learning R

R is very popular and widely-used; in the spirit of the open-source movement many working
statisticians and application scientists have written documentation.

• Introductions and tutorials

• On-line help (within R and on the Internet)

• Contributed documentation

• Textbooks

• Task views

• R Journal, Mailing lists, user’s conference

D G Rossiter
Introduction to R 103

General introductions

• Venables, W. N. ; Smith, D. M. ; R Development Core Team, 2014. An

Introduction to R (Notes on R: A Programming Environment for Data Analysis and
Graphics), updated at each version of R
http://www.cran.r-project.org; also included with R distribution
The standard introduction. This links to:

• Hornik, K. 2007. R FAQ: Frequently Asked Questions on R. Also updated with each
version.
What is R? Why ‘R’ ? Availability, machines, legality, documentation, mailing lists . . .
These are updated with each R release.

D G Rossiter
Introduction to R 104

On-line help

• Within the R environment: help method, abbreviated ?; help.search method

• On the internet

– RSeek: http://www.rseek.org/
– RSiteSearch method

D G Rossiter
Introduction to R 105

RSeek results

D G Rossiter
Introduction to R 106

RSiteSearch method results

D G Rossiter
Introduction to R 107

Textbooks using R
More and more texts are using R code to illustrate their statistical analyses.

• Dalgaard, P. 2002. Introductory Statistics with R. Springer Verlag.

This is a clearly-written introduction to statistics, using R in all examples.

• Venables, W. N. & Ripley, B. D. 2002. Modern applied statistics with S. New

York: Springer-Verlag, 4th edition; http://www.stats.ox.ac.uk/pub/MASS4/
Presents a wide variety of up-to-date statistical methods (including spatial statistics)
with algorithms coded in S; includes an introduction to R, R programming, and R
graphics.

• Fox, J. 2002. An R and S-PLUS Companion to Applied Regression. Newbury Park:

Sage.
A social scientist explains how to use R for regression analysis, including advanced
techniques; this is a companion to his text: Fox, J. 1997. Applied regression, linear
models, and related methods. Newbury Park: Sage

D G Rossiter
Introduction to R 108

The UseR! series

Springer is publishing a series of practical introductions with R code to topics such as:

• data manipulation

• Bayesian analysis

• spatial data anlysis

– Bivand, R. S., Pebesma, E. J., & Gómez-Rubio, V 2008. Applied Spatial

Data Analysis with R: Springer; UseR! series. http://www.asdar-book.org/

• time-series

• interactive graphics

List at http://www.springer.com/series/6991

D G Rossiter
Introduction to R 109

Technical Notes using R

I have written a number of technical notes showing how to accomplish some statistical
tasks with R; the full list is at

http://www.css.cornell.edu/faculty/dgr2/tutorials/index.html

These include general data analysis, logistic regression, confusion matrices, co-kriging,
partioning transects, and fitting rational functions.

D G Rossiter
Introduction to R 110

R Task Views
Some applications are covered in so-called Task Views, on-line at
http://cran.r-project.org/web/views/index.html.

These are a summary by a task maintainer of the facilities in R (e.g., which packages and
functions to use) to accomplish certain tasks. Examples:

• Analysis of Spatial Data

http://cran.r-project.org/web/views/Spatial.html

• Multivariate Statistics
http://cran.r-project.org/web/views/Multivariate.html

D G Rossiter
Introduction to R 111

Keeping up with developments in R

R is a dynamic environment, with a large number of dedicated scientists working to
make it both a rich statistical computing environment and a modern programming
language.

Daily new and modified packages added to CRAN; new versions of the R base
appear 2–4x yr-1

• R Journal: about 4x yr-1; http://journal.r-project.org/

News, announcements, tutorials, programmer’s tips, bibliographies

• Journal of Statistical Software; http://www.jstatsoft.org/

(continued . . . )

D G Rossiter
Introduction to R 112
...

• Mailing lists: “Mailing Lists” link at CRAN:

– R-announce: major announcements, e.g. new versions

– R-packages: announcements of new or updated packages
– R-help: discussion about problems using R, and their solutions. The R gurus monitor
this list and reply as necessary. A search through the archives is a good way to see if
your problem was already discussed.

• useR! user’s conference; proceedings on-line; tutorials, workshops, user presentations,

thematic sessions

D G Rossiter
Introduction to R 113

Topics – Part 4

1. Reproducible research and literate programming

2. The Tidyverse

D G Rossiter
Introduction to R 114

Reproducible research and literate programming

Reproducible research: “research papers with accompanying software tools that allow
the reader to directly reproduce the results and employ the computational methods that are
presented in the research paper.”

Literate programming:

• both code and comments in the same document; code is executed and produces the
results seen in the document; no cut-and-paste

• if data changes, document changes (code is the same, results are different!)

See: Rossiter, DG 2012. Technical Note: Literate Data Analysis using the R environment
for statistical computing and the knitr package 26-December-2012, 35 pp;
http://www.css.cornell.edu/faculty/dgr2/_static/files/R_PDF/LDA.pdf

D G Rossiter
Introduction to R 115

The Tidyverse

• “[A]n opinionated collection of R packages designed for data science1

– The “opinion” of Hadley Wickham

– Main packages dplyr, tidyr, readr, stringr, tibble, ggplot2

• Well-explained in the (free) on-line text R for Data Science2

• Defines a syntax for pipes (magrittr package), for sequences of operations without
having to define intermediate workspace objects

• Defines the tibble: “a modern re-imagining of the data frame, keeping what time has
proven to be effective, and throwing out what it has not.”

1 https://www.tidyverse.org
2 https://r4ds.had.co.nz
D G Rossiter
Introduction to R 116

Pipes
Example:
the_data <-
read.csv('/path/to/data/file.csv') %>%
subset(variable_a > x) %>%
transform(variable_c = variable_a/variable_b) %>%
head(100)

• Only one workspace object (the data) is created

• the results of each expression are passed to the next with the pipe operator %>%.

Exposing variables in a dataframe with the %$% operator :

data(iris) # Edgar Anderson's Iris Data, in datasets package
iris %>%
subset(Sepal.Length > mean(Sepal.Length)) %$%
cor(Sepal.Length, Sepal.Width)

See https://magrittr.tidyverse.org for more examples and complete syntax.

D G Rossiter