R Programming


fname = "Foo"
lname = "Foo"
age   = 42
print(fname)

# This line is a comment


# The next line is some code commented out:
# print(lname)

print(age) # We can also comment here

R and simple math operations

2 + 3     # 5

2 * 3     # 6

2 - 3     # -1

2 / 3     # 0.6666667

2 ^ 3     # 8

R variables

width = 8
height = 3

area_of_recangle = height * width

print(area_of_recangle)  # 24

R assignment (left-assignment, right-assignment)

Both = and <- can be used for left-assignment
-> Can be used for right-assignment

x = 1
y <- 2
3 -> z
print(x)  # 1
print(y)  # 2
print(z)  # 3

Variable types (numeric, character, logical, function)

class
isa
numeric
character
logical
function
numeric: 1 34 2.7
character: "hello" "23"
logical: TRUE FALSE
function: class print

class(23)       # numeric
class(2.3)      # numeric
class(NaN)      # numeric

class("Hello")  # character
class("23")     # character

class(TRUE)     # logical
class(FALSE)    # logical
class(T)        # logical
class(F)        # logical
class(NA)       # logical

class(class)    # function
class(print)    # function

isa(2, "numeric")  # TRUE
isa("2", "character")  # TRUE
isa(FALSE, "logical")  # TRUE

Variable types are deducted

class
isa
numeric
character
logical

x = 23
class(x)     # numeric

x = "George"
class(x)     # character

x = T
class(x)     # logical

isa(2, "numeric")  # TRUE
isa("2", "character")  # TRUE
isa(FALSE, "logical")  # TRUE

paste (join) strings and numbers together

paste

paste("Hello", "World", "!")     # "Hello World !"

name = "Foo"
age = 42
city  = "Budapest"
paste(name, age, city)           # "Foo 42 Budapest"

paste(name, age, city, sep="_")  # "Foo_42_Budapest"

Operator preference order and parentheses

Operators comparing numbers

x = 2
y = 3
z = 3

x > y    # FALSE
x < y    # TRUE
x <= y   # TRUE
x >= y   # FALSE
x == y   # FALSE
x != y   # TRUE

y == z   # TRUE

Operators comparing strings (characters)

characters are compared in ABC order (but not ASCII order!)

"a" < "b"       # TRUE
"a" < "A"       # TRUE
"abc" < "abd"   # TRUE

2 < 11          # TRUE
"2" < "11"      # FALSE


2 < "a"         # TRUE

Convert string (character) to numeric

as.numeric

x = "2"
class(x)           "character"

y = as.numeric(x)
class(y)           "numeric"

Boolean (logical) operations

TRUE & TRUE     # TRUE
TRUE & FALSE    # FALSE

TRUE | FALSE    # TRUE
FALSE | FALSE   # FALSE

! TRUE          # FALSE
! FALSE         # TRUE



# Logical Operations
a = TRUE
b = FALSE

a & b  # logical AND    FALSE
a | b  # logical OR     TRUE
! a    # logical NOT    FALSE

x = 2
y = 4
x & y   # TRUE
x & 0   # FALSE

print(2 & 4)  # TRUE
print(2 & 0)  # FALSE
print(2 | 0)  # TRUE

Concatenate strings

x = "Foo"
y  = "Bar"

# x + y
# Error in x + y : non-numeric argument to binary operator
# Execution halted

z = paste(x, y, sep="")
z     # FooBar
length(z) # 1

# Join elements of a vector
fruits = c("Apple", "Banana")
q = paste(fruits, collapse ="-")
q     # Apple-Banana

# join elements of a numeric vector
numbers = c(2, 3, 4)
class(numbers)  # numeric
nums = paste(numbers, collapse ="-")
nums     # 2-3-4

Convert between types using as.

as.factor

Printing with cat

cat("Hello", "World", "!", end="\n")

Vectors

R vectors

c
c stands for concatenate

x = c(2, 5, 1, 19)
print(x)

y = x + 2
print(y)


z = x * 2
print(z)


lg = log(x)
print(lg)

{% embed include file="src/examples/vectors/vector.out)

Variable types of vectors are deducted

We'll learn a lot more about vectors later
TODO: Mixed data?

y = c(2, 7, 3)
class(y)     # numeric

the_truth = c(TRUE, FALSE, TRUE, TRUE, FALSE)
class(the_truth)  # logical
length(the_truth) # 5

One element vector is the same a single value

name = "foo"
names = c("foo")
print(name == names)
class(name)
class(names)
print(name[1])
print(names[1])

{% embed include file="src/examples/vectors/one_element_vector.out)

Sum of values in vectors

y = c(2, 7, 3)
sum(y)       # 12

sum(TRUE)    # 1
sum(FALSE)   # 0

the_truth = c(TRUE, FALSE, TRUE, TRUE, FALSE)
sum(the_truth)    # 3

Size of vector (length of vector)

distances = c(11, 12, 13, 14)

length(distances)   # 4

Access the n-the element of a vector

It is some_vector[n]
It is 1-based

distances = c(11, 12, 13, 14)

distances[1]        # 11
distances[4]        # 14
distances[5]        # NA

R Vector: Negative index, exclude element

Using a negative index will give you the same vector without that element

distances = c(11, 12, 13, 14)

nd = distances[-2]
print(nd)           # 11 13 14
print(distances)    # 11 12 13 14

R vector: Access several elements (slice)

We can use a vector as the index of anther vector thereby fetching a new vector of the specific values.
We can also select a range of elements using :

distances = c(11, 12, 13, 14, 15, 16, 17, 18)

print(distances[c(2, 4)])  # 12, 14
print(distances[2:4])  # 12, 13, 14


locations = c(5, 3, 5, 7)
nd = distances[locations]

print(nd)            # 15, 13, 15, 17
print(distances)

R vector: Exclude several elements

distances = c(11, 12, 13, 14, 15, 16, 17, 18)

print(distances[-c(2, 4)])  # 11 13 15 16 17 18
print(distances[-(2:4)])    # 11 15 16 17 18

Some basic statistical functions

median
sum
mean
min
max

distances = c(11, 12, 13, 14, 15, 16, 17, 180, 7, 11)

sum(distances)
median(distances)
mean(distances)   # average
min(distances)
max(distances)
var(distances)    # variance
st(distances)     # standard deviation

3 ways to create vectors

x = 1:3
y = seq(1, 3, 1)
z = c(1, 2, 3)
print(x)
print(y)
print(z)
print(x == y)
print(x == z)
print(y == z)

# 1 2 3
# 1 2 3
# 1 2 3
# TRUE TRUE TRUE
# TRUE TRUE TRUE
# TRUE TRUE TRUE

R - sequences and ranges

nums = seq(1, 10, 2)
print(nums)  # 1 3 5 7 9
print(class(nums))  # "numeric"
print(is.numeric(nums))  # TRUE

other = seq(to = 19, from = 7, by = 3)
print(other)  #  7 10 13 16 19

Range of numbers

range_of_numbers = 1:10
print(range_of_numbers) #  1  2  3  4  5  6  7  8  9 10

Filter values

values = c(17, 3, 28, 5)
large_logic = values > 10
large = values[large_logic]
print(large)    # 17, 28


n = c(2, 4, 7, 6, 9, 8)
n[  n > 4 ]
n[  n > 4 & n %% 2] # grater than for and non-zero modulus 2
# n[ some-logical-vector ] to filter certain values and create a shorter vector
n[  n >= 4 ]

Index of true elements - which

which

# the index of the elements that have TRUE value
which(c(TRUE, FALSE, FALSE, TRUE))  # 1, 4

which

Showing the indexes of the elements that are TRUE in a boolean vector

n = c(2, 4, 7, 6, 9, 8)
which(n > 4)
which(n > 4 & n %% 2)

Matix TBD

# Show a matrix

Vector operations - reuse values from shorter vector

x = c(1, 2, 3, 4)
y = c(10, 20)
x+y

# There is no warning here !

x = c(1, 2, 3, 4)
y = c(10, 20)
print(x*y)  #  10 40 30 80
# 1 * 10
# 2 * 20
# 3 * 10
# 4 * 20

# R starts to reuse the shorter vector if one of the vactors is longer than the other
# gives a warning
# TODO: can we stop running at such warnings?

a = c(TRUE, TRUE, FALSE)
b = c(TRUE, FALSE, TRUE, FALSE)

a & b
a | b
! a

x = c(1, 2, 3)
y = c(10, 20)
x+y

# options(warn=2)  # turn warnings into errors
# options(warn=0)  # ???? turn warnings back to warnings

Summary of numeric data

summary
mean

nums = c(2, 4, 5)
sum(nums)
mean(nums)
max(nums)
min(nums)
median(nums)
summary(nums)

bools = c(T, F, T, T, F)
summary(bools)
mean(bools)
max(bools)
min(bools)
sum(bools)

names = c("biology", "chemistry", "physics")
summary(names)
#mean(names)
max(names)
min(names)
#sum(names)

Change element of vector

colors = c("red", "blue", "green")
colors    # "red", "blue", "green"
colors[1] # red
colors[1] = "purple"
colors # "purple", "blue", "green"

Assign vector to another name

colors = c("red", "blue", "green")
colors    # "red", "blue", "green"

other_colors = colors

colors[1] # red
colors[1] = "purple"
colors # "purple", "blue", "green"
other_colors  # "purple", "blue", "green"

Reverse vector

rev
reverse

fruits = c("apple", "peach", "lemon", "banana", "ananas")
rev(fruits)  # "ananas" "banana" "lemon"  "peach"  "apple"

Sort vector

sort

numbers = c(7, 3, 9, 11, 2)
sort(numbers)  # 2  3  7  9 11

fruits = c("apple", "peach", "lemon", "banana", "ananas")
sort(fruits)                   # "ananas" "apple"  "banana" "lemon"  "peach" 
sort(fruits, decreasing=TRUE)  # "peach"  "lemon"  "banana" "apple"  "ananas"

Sort using order

order

numbers = c(7, 3, 9, 11, 2)
order(numbers)   # 5 2 1 3 4
numbers[order(numbers)]  # 2  3  7  9 11

fruits = c("apple", "peach", "lemon", "banana", "ananas")
order(fruits)                 #  5 1 4 3 2
fruits[ order(fruits) ]       # "ananas" "apple"  "banana" "lemon"  "peach"
fruits[ order(fruits, decreasing = T) ]  # "peach"  "lemon"  "banana" "apple"  "ananas"

Operators comparing vectors

# Show how to compare a vector with an individual value (=> this just plays on the shorter being repeated to the length of the longer)
x = c(1, 2)
y = c(11, 0)

x > y
x < y
x <= y
x >= y
x == y
x != y

Convert vector of strings to numerics

as.numeric

chars = c("1", "2", "3")
nums = as.numeric(chars)

print(class(chars))
print(class(nums))

print(summary(chars))
print(summary(nums))
print(sum(nums))   # 6
print(sum(chars))  # invalid 'type' (character) of argument

Repeate the same number

same_number = rep(2, 10)
print(same_number) # 2 2 2 2 2 2 2 2 2 2

Boolean (logical) operations on vectors

a = c(TRUE, TRUE, FALSE, FALSE)
b = c(TRUE, FALSE, TRUE, FALSE)
sum(a)
sum(b)  # number of TRUE items

a & b
a | b
! a

a = c(TRUE, TRUE, FALSE, TRUE)
b = c(TRUE, FALSE, TRUE, FALSE)
c = c(TRUE, FALSE, FALSE, TRUE)

a & b
a | b
! a

# operator precedence
# the use of parentheses

a & (b | c)

Factors

"Category", "enumerated type"

m = c("apple", "apple", "banana", "apple", "peach", "banana", "apple")
m
f = as.factor(m)

summary(m)
summary(f)
levels(f)
# levels - possible values a variable iny can have

fruits_vector = c("Apple", "Banana", "Apple", "Apple", "Peach", "Banana")
fruits_factor  = as.factor(fruits_vector)

fruits_vector   # [1] "Apple"  "Banana" "Apple"  "Apple"  "Peach"  "Banana"

fruits_factor   # [1] Apple  Banana Apple  Apple  Peach  Banana
                # Levels: Apple Banana Peach

class(fruits_vector)  # "character"
class(fruits_factor)  # "factor"

length(fruits_vector) # 6
length(fruits_factor) # 6

levels(fruits_vector) # NULL
levels(fruits_factor) # "Apple"  "Banana" "Peach"

fruits_vector["Apple"] # NA
fruits_factor["Apple"] # [1] <NA>
                       # Levels: Apple Banana Peach

Append to end of vector

animals = c("cat", "dog")
length(animals)

animals = append(animals, "mouse")
length(animals)

Data Frames

Data Frame functions

data.frames dim head nrow ncol names - get the names or assign new names to the data.frame names(dataframe) = c("title1", "title2", ...) summary Subsetting (indexing ranges) subset() order() returns a numeric vector of the sorted indexes

Iris dataset

Built-in Iris data set

library(datasets)
#write.table(file="a.txt", iris, sep="\t")
iris

filename = "/home/gabor/Dropbox/Weizmann/R/iris_dataset.txt"

# setwd - set working directory
# getcwd - get working directory
ir = read.table(filename, sep="\t", header=T)
head(ir, 3)
tail(ir, 3)
View(ir)   # Capital V !!! (opens a separate view of the data)
class(ir)  # data.frame
dim(ir)  # 150 5  dimensions: (rows, columns)
ir[2, 3]
head(ir)
# The indexes on the left hand side are also called "row-names"

ir[,1]     # access a column by its location
ir$seplen  # access a column by its name 

ir1 = ir[, c("seplen", "sepwid")]  # dataframe
head(ir1)

column = ir[, c("seplen")] # numberic vector

summary(ir)

Load Iris dataset

iris = read.table(file="data/iris.txt", sep=" ", header=T, stringsAsFactors=T)
head(iris)

# hist(iris$Sepal.Length)
hist(iris$Petal.Width)

mean_petal_width = mean(iris$Petal.Width)
hist(
    x=iris$Petal.Width,
    breaks=10,
    col="light blue",
    main="Distribution of Petal Width",
    xlab="Width of Petal (cm)",
    #ylab="",
    sub=paste("Mean", mean_petal_width), # paste concatentes values
    )

#help(hist)
#View(iris)
#tail(iris)
# iris

#class(iris)
#dim(iris)   # rows, columns

# iris[2, 4]
# iris[1, ]
# iris[, 1]
#iris$Sepal.Length
#iris['Sepal.Length']   # numeric vector
#iris[c('Sepal.Length', 'Petal.Length')]  # data.frame
#iris[,c('Sepal.Length', 'Petal.Length')] # data.frame
#summary(iris)
colnames(iris)


plot(iris$Petal.Length, iris$Sepal.Length)
help(plot)  # or ?plot

plot(
  x=iris$Petal.Length,
  y=iris$Sepal.Length,
  #type="l",  # or p for points
  #pch=20,  # change how the points look like

  # col (short for color)
  #col="purple red",

  col=iris$Species,
  #pch=19

  xlab="X Title",
  ylab="Y Title",
  main="Main heading",
  sub="Sub heading",
)

# In the iris dataset when we draw a histogram how does it pick colors when we only have names?
# Because it is a factor (beacise of the stringsAsFactors)
class(iris$Species) # "factor"
# Each "level" in the factor has a numerical value (1, 2, 3) and in R each color also a number (1 = black 2 = red, etc)


plot(
  x=iris$Petal.Length,
  y=iris$Sepal.Length,
  pch=as.numeric(iris$Species),
)

# TODO: how to pick the specific colors?

# how to color the point accorind to some other condition? e.g. iris$Sepal.width > 2
plot(
  x=iris$Petal.Length,
  y=iris$Sepal.Length,
  col=as.numeric(iris$Sepal.Width > 3)+1
)
# iris$Sepal.Width > 2   is a boolean vector
# as.numeric(iris$Sepal.Width > 2) is a numerical vector of 0 and 1 values
# careful: color 0 is white so we won't see it, that's why we add 1 so instead of 0 and 1 we will get 1 and 2 values.


levels(iris$Species) # "setosa" "versicolor" "virginica"

help(data.frame)


pairs(iris[,1:4])  # pairwise relation graphs
pairs(iris[,1:4], col=as.numeric(iris$Species))
pairs(iris[,1:4], col=as.numeric(iris$Species), upper.panel=NULL)

Sort dataset by a column

iris = read.table(file="data/iris.txt", sep=" ", header=T, stringsAsFactors=T)
head(iris)
ord = order(iris$Sepal.Length)
head(iris[ord,])

Order dataframe by two columns (secondary sort)

df = data.frame(A=c(2, 1, 1, 3), B=c(2, 4, 3, 1))
df
df[order(df$A),]
df[order(df$B),]
df[order(df$A, df$B),]

Strings (characters)

Substitute first occurence using sub

text <- "One cat two cats and another single cat"
print(text)
print(sub('cat', 'dog', text))
print(text)

[1] "One cat two cats and another single cat"
[1] "One dog two cats and another single cat"
[1] "One cat two cats and another single cat"

Substitute all occurences using gsub

gsub

text <- "One cat two cats and another single cat"
print(text)
print(gsub('cat', 'dog', text))
print(text)

[1] "One cat two cats and another single cat"
[1] "One dog two dogs and another single dog"
[1] "One cat two cats and another single cat"

Functions

Functions overview

function_name = function(arg1, arg2, arg3) {
     # commands
     return(someReturnValue)
}

Function parameter names are local to the function
variables created in the function are local
Result of last expression returned implicitely (even without calling return())
When accepting a vector of numbers what if the user sends in a vector of strings (different class of data)?
What if some of the values are missing? (NA)
What if the vector is empty?

Simple add function

rectangle_area=function(a, b) {
  area = a*b
  return(area)
}
rectangle_area(2, 3)
rectangle_area(4, 7)
# area - not defined here

a = c(0, 0)
b = c(3, 4)
distance=function(a, b) {

  xdiff = abs(a[1]-b[0])
  print(xdiff)
  ydiff = abs(a[1]-b[1])
  print((xdiff**2 + ydiff**2)**0.5)
}
distance(a, b)

Recursive Fibonacci

fibonacci=function(n) {
    if (n == 1) {
        return(1)
    }
    if (n == 2) {
        return(1)
    }
    return(fibonacci(n-1) + fibonacci(n-2))
}

for (n in 1:10) {
    cat(fibonacci(n), end="\n")
}

Fibonacci

fibonacci=function(n) {
    if (n == 1) {
        return(1)
    }
    if (n == 2) {
        return(1)
    }
    fibs = c(1, 1)
    for (i in 3:n) {
        last = fibs[length(fibs)] + fibs[length(fibs)-1]
        fibs = append(fibs, last)
    }
    return(fibs)
}

cat(fibonacci(10), end="\n")

Testing

Testthat

.libPaths(append(.libPaths(), "lib"))
install.packages("testthat", "lib")
library("testhat")

Test Example

add=function(x, y) {
    return(x+y)
}

multiply=function(x, y) {
    return(x*y)
}

Using the functions

.libPaths(append(.libPaths(), "lib"))
install.packages(setdiff("this.path", rownames(installed.packages())), "lib")
library("this.path")
root = dirname(this.path())
source(file.path(root, "mymath.R"))

if (add(2, 3) != 5) {
    exit(1)
}

if (add(2, -2) != 0) {
    exit(1)
}

Files

Filesystem pathes

dirname
basename


basename("/home/foobar/projects/my_r_project/main.R") # "main.R"
dirname("/home/foobar/projects/my_r_project/main.R")  # "/home/foobar/projects/my_r_project"

Filesystem get current working directory (cwd)

getwd
cwd

getwd()      # "/home/gabor/work/slides/r"

Concatenate pathes

file.path

file.path("usr", "local", "lib")  # "usr/local/lib"

Change directory

getwd
setwd

getwd()
setwd("examples")   # change directory
getwd()
setwd("/home/gabor/work/slides/r")
getwd()

Recursively list files and directories

list.dirs
list.files

list.dirs(".", recursive=FALSE)  # recursive defaults to TRUE

list.files(".", recursive=TRUE)  # recursive defaults to FALSE

Write to a textfile

file
cat

filename = "sample.txt"
print(filename)

fh <- file(filename, "w")
cat("TITLE line", "First row", "", "Third row", sep = "\n", file = fh)
cat("One more line\n", file = fh)
close(fh)

Read a textfile

filename = "README.md"

readLines(filename)

Counter with file storage

filename = "count.txt"
count <- 0

if (file.exists(filename)) {
    count <- as.numeric(readLines(filename))
}

count <- count + 1

cat(count, end="\n")

fh <- file(filename, "w")
cat(count, end="\n", file = fh)
close(fh)

Graphs

Bar plot

barplot

distances = c(11, 15, 7, 23, 9)
barplot(distances)

barplot(c(2, 3, 7, 1))
data <- rnorm(100)
barplot(data)

barplot(c(2, 3, 7, 1))

data <- rnorm(100)
barplot(data)

Associate name with each value

fruits = c("Apple", "Banana", "Peach")
nums = c(3, 7, 5)

names(nums) = fruits

print(nums["Peach"])
#print(class(nums["Peach"]))

{% embed include file="src/examples/graphs/associate_name.out)

barplot this!

Quick graphs

barplot
plot
hist
boxplot
pie

distances = c(11, 15, 7, 23, 9)
barplot(distances)
plot(distances)
hist(distances)
boxplot(distances)
pie(distances)


barplot(distances, col="purple")
barplot(distances, col="#2323AA")
barplot(distances, col="#2323AA", ylab="Text on Y", xlab="Text on X")

Will create a file called Rplots.pdf.

Random numbers in normal distribution

rnorm
hist
Normal distribution
Histogram of the numbers

normal_numbers = rnorm(100)
hist(normal_numbers)

Packages

CRAN

CRAN - The Comprehensive R Archive Network

Install CRAN packages

$ R
> install.packages("csv", "lib")

This will install the "csv" package in the "lib" folder. It will also install all the dependencies.

Install CRAN packages from the command line

Rscript -e 'install.packages("RUnit")'

Install several CRAN packages from the command line

Rscript -e 'install.packages(c("RUnit", "zoo"))'

Read CSV file


.libPaths(append(.libPaths(), "lib"))
install.packages("csv", "lib")
library("csv")

filename = "a.csv"
table = read.table(filename)

Read JSON file

JSON
read_json
jsonlite

.libPaths(append(.libPaths(), "lib"))
install.packages("jsonlite", "lib")
library("jsonlite")

filename = "r.json"
data = read_json(filename)

Other

Exit - quit

quit

![](examples/other/exit.R" %}

Script name

![](examples/other/script_name.R" %}

Command line arguments

![](examples/other/command_line_arguments.R" %}

if-statement

if
else

![](examples/other/if.R" %}

for loop

![](examples/other/for.R" %}

Keyboard shortcuts

R Programming