R

1
R

• Basics
• Based on sheets provided by
• prof. dr. Han van der Maas (UvA)

2
Part 1 What is R?

• a programming environment
• powerful calculator
• excellent graphics capabilities
• excellent statistical procedures (advanced SPSS)
• Replaces C, Excel and SPSS
• Computer simulations
• Data management
• Statistical analysis
• Figures for papers, talks etc.

3
Why R?

• easy to learn
• powerful
• Freeware http//cran.r-project.org/
• platform independent
• supported by a large user network
• R has a very successful package system for
  distributing code and data
• can be used in combination with C, Fortran
• object-oriented

4
Disadvantages of R

• some libraries are not available in R
• R (and Matlab) can not be used for
  experimentation (Authorware Presentation)
• R is slow and uses lots of memory
• But computers are fast and have lots of memory
• In extreme cases we combine R and C

5
What about Matlab?

• Matlab is very expensive (student version already
  hundreds of dollars)
• R and Matlab are very similar, but some things
  are easier in Matlab, and other things are easier
  in R
• Psychologists will learn Matlab

6
Demo 1

• 34
• a lt- 3
• a4
• a lt- c(3,1,2,5,6, 8,6,9,11)
• a4
• B lt- matrix(a,3,3)
• B4

• matplot(B,type'b')
• C lt- matrix( rnorm(200),40,5)
• matplot(C,type'b')
• pairs(C)
• t.test(C)
• cor(C,1,C,2)

7
Demo 2

• data
• xseq(-5,3,length100)y23xx2-2x3rnorm(100
  ,0,20)
• fit models
• m1lm(yx)
• m2lm(ypoly(x,2))
• m3lm(ypoly(x,3))
• m4lm(ypoly(x,4))
• summary(m3) coefficients and statistics
• summary(lm(yxI(x2)I(x3)))
• compare models
• anova(m2,m3)anova(m3,m4)
• compute predicted dvalues
• pred1predict(m1,interval"conf",newdatadata.fram
  e(x))
• pred2predict(m2,interval"conf",newdatadata.fram
  e(x))
• pred3predict(m3,interval"conf",newdatadata.fram
  e(x))
• pred4predict(m4,interval"conf",newdatadata.fram
  e(x))
• and plot
• plot(x,y)matlines(x,pred3,ltyc(1,2,2))
• lines(x,pred1,1,lty3) lines(x,pred2,1,lty4)
  

8
Demo 3
o p t i o n a l

• help.start()
• Choose packages
• Try nnet
• Copy example
• Go to R
• library(nnet)
• Paste example
• You have run a neural network simulation !

9
Learning strategy

• Lectures
• Main texts Introduction to R
• Lots of computer exercises
• Lots of work
• You will really learn it only if you use it in
  all your subsequent work!
• Have patience.

10
Overview

• Meeting 1 Basics
• Meeting 2 Data
• Meeting 3 Functions
• Meeting 4 Programming I
• Meeting 5 Programming II
• Meeting 6 Graphics
• Meeting 7 Statistical
• Meeting 8 Extra
• Meeting 9 Exam on October 28 10.00h-17.00h
• Excercises work alone, work hard.

11
Using R

• R is a scripting language
• Read, evaluate, print result
• Write your code in an .R file
• Tabs File, Open script
• You can also save the history of your commands
• Use arrow keys to scroll to previous commands
• Use for comments

12
Help!

• There are many ways to get help
• help(lm) or ?lm (check examples)
• apropos(lm)
• help.start() to open browser
• On the internet
• http//cran.r-project.org/
• Search
• R manuals
• R wiki
• R reference card

13
R basics

• Basically two types of objects
• Data (variables, vectors, matrices, lists)
• Functions (operators, plot, statistical,
  mathematical functions)
• In R there are hundreds of functions available
  (and you can make new ones yourself)
• functions always have the form f(), e.g.,
  mean(x), plot(x,y)
• This is also true for operators like , you can
  write 34, but this is short hand for ''(3,4)

14
Data

• Information
• Variables
• Matrices with numbers
• Datasets
• Strings
• ...
• For example
• alt-2
• Then a is data object

15
Workspace

• Data exist in the workspace (active memory)
• With ls() you can see your objects
• With rm() you can remove them

16
Assignment (function)

• a 1
• a lt- 1
• 1 -gt a
• assign(a,1)
• a b 1 or a lt- b lt- 1
• Note test equality with (or all.equal())
• Separate commands with a newline or

17
Operators (function)

• - Minus, can be unary or binary
• Plus, can be unary or binary
• ! Unary not
• Tilde, used for model formulae, can be either
  unary or binary
• ? Help
• Sequence, binary (in model formulae
  interaction)
• Multiplication, binary
• / Division, binary
• Exponentiation, binary
• any Special binary operators, any can be
  replaced by any valid name
• Modulus, binary
• / Integer divide, binary
• Matrix product, binary

• o Outer product, binary
• x Kronecker product, binary
• in Matching operator, binary (in model
  formulae nesting)
• lt Less than, binary
• gt Greater than, binary
• Equal to, binary
• ! Unequal to, binary
• gt Greater than or equal to, binary
• lt Less than or equal to, binary
• And, binary, vectorized
• And, binary, not vectorized
• Or, binary, vectorized
• Or, binary, not vectorized
• lt- Left assignment, binary
• Left assignment, binary
• -gt Right assignment, binary
• List subset, binary

18
Precedence of operators(from highest to lowest)
o p t i o n a l

• ( Indexing
• access variables in a name space
• _at_ component
• / slot extraction
• exponentiation
• - unary minus and plus
• sequence operator
• any special operators
• / multiply, divide
• - (binary) add, subtract

• lt gt lt gt ! ordering and comparison
• ! negation
• and
• or
• as in formulae
• -gt -gtgt assignment
• assignment
• lt- ltlt- assignment
• ? help

19
Simple math

• x lt- .5
• sin(x) cos(x) tan(x)
• asin(x) acos(x) atan(x)
• log(x) log10(x) exp(x) sqrt(x)
• xx
• Use parentheses to be sure
• -22 is how much?
• And 2-2 ?

20
So you can do things like...

• a lt- 3 b lt- 5
• y lt- sqrt(a2b2)
• !(alt1blt1)
• But there is much more...
• More complex variables
• More interesting operators (functions)
• A lot of functions are always available (base
  package), others can be added or created