• a free software for data analysis
• an interpreted programming language, derived from `S-plus’
• initially developed by R. Ihaka and R. Gentleman (1996)
• currently developed by the R Core Team (20 people)
• largest collection of tools for data analysis (1,000s of contributors)

• The R project: www.r-project.org
• archiving / distribution network CRAN:
  cran.r-project.org/mirrors.html
• available on Windows, MacOSX, Linux

• basic statistics: statistical tests, linear modelling, multivariate analysis
• spatial statistics: GIS, mapping, clustering
• graph theory: social sciences, network analysis, graph algorithm
• genetics: phylogenetic trees, genetic markers, genomics
• epidemiology!
• much more: see task views cran.r-project.org/web/views/

• Freedom = ability to make informed decisions
• you don’t pay for it
• the code is accessible by anyone
• anyone can use, modify and share the code

• get R for your system (download from CRAN)

• get a Graphical User Interface (GUI): RStudio, emacs + ESS, Tinn-R

• (or at least) get a text editor to write code: notepad++, emacs, vi, Tinn-R, …

• ?foo or help("foo"): access the help page of foo
• ??bar or help.search("foo"): look for foo in help pages
• dedicated mailing lists: stat.ethz.ch/mailman/listinfo
• the RECON forum: http://www.repidemicsconsortium.org/forum/
• google

• no files, all in the RAM (i.e. temporary memory)
• data, results, functions, etc. are all R objects
• one object can be saved / loaded using saveRDS/readRDS (output: .rds files)
• several objects can be saved / loaded using save/load (output: .RData files)
• an entire session can be saved using save.image

General syntax: object_name <- content:

toto <- 1:8
toto # check content: 1, 2, 3, ...

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

toto <- "some text"
toto # content has changed

## [1] "some text"

a <- 1:10
a

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

class(a)

## [1] "integer"

b <- c(-0.1, 10.123, pi)
b

## [1] -0.100000 10.123000  3.141593

class(b)

## [1] "numeric"

a <- c("hello world", "this is fun", "even more fun coming")
a

## [1] "hello world"          "this is fun"          "even more fun coming"

class(a)

## [1] "character"

a <- factor(c("red", "blue", "green", "red", "green"))
a

## [1] red   blue  green red   green
## Levels: blue green red

class(a)

## [1] "factor"

levels(a)

## [1] "blue"  "green" "red"

The logical type can be TRUE or FALSE:

a <- c(TRUE, FALSE, TRUE, TRUE)
a

## [1]  TRUE FALSE  TRUE  TRUE

class(a)

## [1] "logical"

A vector stores several values of the same type as a one-dimensional array:

a <- c(1, 2, 10, -1, 1.123)
a

## [1]  1.000  2.000 10.000 -1.000  1.123

length(a)

## [1] 5

A matrix stores several values of the same type as a table:

a <- matrix(sample(1:12), ncol = 4)
a

##      [,1] [,2] [,3] [,4]
## [1,]   12    7    8    4
## [2,]   11   10    2    5
## [3,]    1    6    9    3

class(a)

## [1] "matrix"

dim(a)

## [1] 3 4

A data.frame is a table where variables (columns) can have different types (equivalent to a spreadsheet):

a <- data.frame(age = c(10, 54, 3), sex = c("m", "f", "m"))
a

##   age sex
## 1  10   m
## 2  54   f
## 3   3   m

class(a)

## [1] "data.frame"

dim(a)

## [1] 3 2

A list is a collection of objects of any types and sizes, stored as different slots:

age <- c(10, 54, 3)
sex <- factor(c("m", "f", "m"))
swab  <- matrix(
  sample(c("+", "-"), replace = TRUE, 10), nrow = 2,
  dimnames = list(NULL, paste("t", 1:5, sep = "")))
x <- list(age = age, gender = sex, swab_results = swab)

x

## $age
## [1] 10 54  3
## 
## $gender
## [1] m f m
## Levels: f m
## 
## $swab_results
##      t1  t2  t3  t4  t5 
## [1,] "+" "+" "-" "+" "+"
## [2,] "+" "+" "+" "-" "+"

class(x)

## [1] "list"

length(x)

## [1] 3

• integer: integer numbers
• numeric: decimal numbers
• character: character strings
• factor: categorical variables
• vector: collection of values (1 dimensional array, same type)
• matrix: collection of values (table, same type)
• data.frame: columns can have different types, but same length (spreadsheet)
• list: collection of elements, no restriction of content
• ...: classes for: DNA sequences, maps, networks, etc.

A set of operations made on a given input.

Syntax: function_name(argument1, argument2, ...)

Example:

rnorm(8, mean = 5, sd = 3)

## [1]  2.713883  7.373878  9.364017  1.173302  6.436040  5.624428  7.107626
## [8] -1.727200

Read the documentation by typing ?function_name.

Example:

?rnorm

For example:

rnorm(5, mean = 3, sd = 2) # this result is not stored

## [1] -1.133996  7.153785  3.201488  1.897321  1.427742

toto <- rnorm(5, mean = 3, sd = 2) # store output in toto
toto

## [1] -0.04287389  7.92735044  3.46023461  0.21262429  3.52734135

For example:

hist(rnorm(1000, mean = 3, sd = 2), col = rainbow(15))

Objects can be subsetted by index, name, or logical, using:

• object_name[] for a vector
• object_name[rows, columns] for a matrix / data.frame
• object_name[[]] for a list

x[foo] where foo is an integer vector

• positive integers: position of retained entries
• negative integers: position of discarded entries

x <- 10:1
x

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

x[c(1, 2, 5)]

## [1] 10  9  6

letters[2:10]

## [1] "b" "c" "d" "e" "f" "g" "h" "i" "j"

letters[-(1:20)]

## [1] "u" "v" "w" "x" "y" "z"

x[foo] where foo is a logical vector, returns the values of x where foo is TRUE:

x <- 1:10
x < 5

##  [1]  TRUE  TRUE  TRUE  TRUE FALSE FALSE FALSE FALSE FALSE FALSE

x[x < 5]

## [1] 1 2 3 4

x[foo] where x is named and foo is a character vector of retained names:

x <- sample(1:100, 10)
names(x) <- letters[1:10]
x

##   a   b   c   d   e   f   g   h   i   j 
##  92 100   8  53   7  22  35  49   2  71

x[c("c", "d", "a", "i")]

##  c  d  a  i 
##  8 53 92  2

x[foo] <- new.value where x[foo] are values to be replaced with new.value:

x <- round(rnorm(14), 2)
x

##  [1] -1.32  0.12 -0.78  0.68  1.52 -0.06  0.95 -1.97  1.33 -1.42 -0.40
## [12]  2.35 -0.05  0.88

x[x< 0] <- 0
x

##  [1] 0.00 0.12 0.00 0.68 1.52 0.00 0.95 0.00 1.33 0.00 0.00 2.35 0.00 0.88

Same principle, with x[lines, columns]:

x <- matrix(1:15, nrow = 3)
x

##      [,1] [,2] [,3] [,4] [,5]
## [1,]    1    4    7   10   13
## [2,]    2    5    8   11   14
## [3,]    3    6    9   12   15

x[c(3,1), ]

##      [,1] [,2] [,3] [,4] [,5]
## [1,]    3    6    9   12   15
## [2,]    1    4    7   10   13

x[2, 4:5]

## [1] 11 14

x[foo] to return a list, x[[foo]] for a single element:

x <- list(a = rnorm(4), hi = "Hello", stuff = letters[1:10]); x

## $a
## [1] -0.03960258  1.20769218 -0.64288920  1.63578167
## 
## $hi
## [1] "Hello"
## 
## $stuff
##  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"

x[c(1,3)]

## $a
## [1] -0.03960258  1.20769218 -0.64288920  1.63578167
## 
## $stuff
##  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"

x[[3]]

##  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"

x$stuff

##  [1] "a" "b" "c" "d" "e" "f" "g" "h" "i" "j"

a <- c(TRUE, TRUE, FALSE, FALSE)
b <- c(TRUE, FALSE, TRUE, FALSE)
rbind(a,b)

##   [,1]  [,2]  [,3]  [,4]
## a TRUE  TRUE FALSE FALSE
## b TRUE FALSE  TRUE FALSE

a & b # logical 'AND'

## [1]  TRUE FALSE FALSE FALSE

a | b # logical 'OR'

## [1]  TRUE  TRUE  TRUE FALSE

!a # not A

## [1] FALSE FALSE  TRUE  TRUE

any(a) # at least one TRUE

## [1] TRUE

all(a) # all TRUE

## [1] FALSE

which(a) # indices of the TRUEs

## [1] 1 2

is.na is useful to spot missing data (NAs)

a <- c(3,NA,2,5,NA,10)
is.na(a)

## [1] FALSE  TRUE FALSE FALSE  TRUE FALSE

a[!is.na(a)]

## [1]  3  2  5 10

Logicals are numbers: TRUE = 1, FALSE = 0

a <- -10:5
sum(a > 0)

## [1] 5

mean(a > 0)

## [1] 0.3125

a * (a > 0)

##  [1] 0 0 0 0 0 0 0 0 0 0 0 1 2 3 4 5