• R objects can be saved in .RData and .Rds files; .RData files can contain multiple objects.
• Tabular data can be loaded from or saved into delimited text files (.txt, .csv, .xslx, …).
• R functions and scripts can be saved as .R files.
• R packages containing additional functions can be imported from local archive files or repositories like CRAN, Bioconductor, and GitHub.

A path: indicates the unique location of a file or directory.

A path is expressed by a string of characters separated by a delimiting character, where each component separated by the delimiting characters represents a directory.

Absolute path: points to the same location in a file system, regardless of the current working directory. To do that, it must include the root directory.

Relative path: starts from some given working directory. A file name alone can be considered as a relative path based at the current working directory.

Examples:

• /home/Desktop/myscript.R
• ../Data/myfile.txt
• ./myplot.pdf
• myplot.pdf

Let’s suppose that our working directory is:

/Users/francescafinotello/Desktop/R_course_MUI/Code

We can load/save whatever R script in the “Code” directory just by specifying the file name.

But if we want to load/save a file from/in the “Data” directory we need to specify its location:

• Absolute path:
  /Users/francescafinotello/Desktop/R_course_MUI/Data/myfile.txt

• Relative path:
  ../Data/myfile.txt

The getwd function can be used to known the absolute path of your working directory:

getwd()

## [1] "/Users/francescafinotello/Dropbox/R_course_MUI_2021/Course_slides"

The setwd function can be used to set the path (absolute or relative) of the working directory:

setwd("/Users/francescafinotello/Desktop/")

To set the working directory using the toolbar: Session > Set Working Directory

To know which files are present in your working directory (or in another directory) you can use the list.files function:

list.files()
list.files("../")

This is different from asking which objects are present in the R workspace using the ls function:

ls()

To clean your R workspace you can use the rm function or the broom in the Environment panel.

rm( list = ls() )

The function source can be used to

• Run a script (e.g., full analysis or exercises from day 1)

source("RNAseq_diff_gene_expr.R")
source("../Scripts/Day1_Ex8.R")

• Import one or more R functions you want to use (e.g., your BMI function)

source("BMI_function.R")

The function save can be used to save one ore more R objects into an .RData file. The file argument must always be present to specify the path to the output file

age <- 40
height <- 180
save(age, height,                         # List of objects to be saved
  file = "../Data/Day3_data_example.RData")  # Path to output file

The function load can be used to import .RData files

load("../Data/Day3_data_example.RData")

weight <- 80
save(age, height, weight, 
  file = "../Data/Day3_data_example_updated.RData")

Function: set of statements organized together to perform a specific task.

R has a large number of built-in functions.

Additionally, users can:

• Import R packages enclosing a set of functions for specific tasks (e.g., differential gene expression analysis, machine learning)

• Create new functions.

Functions are usually invoked with their name, followed by round brakets listing the arguments to be considered (if any).

v1 <- c(1, 3, 56, 6)
v2 <- c(4, 5, 40, 7)
cor(x=v1, y=v2, method="pearson")

## [1] 0.999966

The name of the arguments can be omitted if the right order of the arguments is respected:

cor(v1, v2, method="pearson")

## [1] 0.999966

So far, we have used already several functions:

log10, sqrt, class, is.na, as.character, as.factor, as.numeric, c, length, names, levels, which, rep, seq, factor, mean.

There are other arithmetical functions:

sum, prod, max, min, which.max, which.min, range, median, var, sd, round, sign, exp.

If we do not know how a function works, we can invoke the help function to understand:

• Which task it performs
• Which are its arguments, their expected order, and their default values
• Which results it generates (e.g. a plot or an object with a specific format)
• If there are similar functions
• How it can be used, explained through examples

help(mean)

help(sample)

x <- c("Schildkroete", 
  "Ohrwurm", 
  "Rechtsschutzversicherungsgesellschaften")

length(x) # Length of the vector of strings

## [1] 3

nchar(x) # Length of the strings

## [1] 12  7 39

x <- c("a", "b", "c", "c", "d", "e", "e", "e")

unique(x) # Unique elements of x

## [1] "a" "b" "c" "d" "e"

table(x) # Occurrences of each value in x

## x
## a b c d e 
## 1 1 2 1 3

x <- c(10, 2, 20, -1)

sort(x) # Sorted x

## [1] -1  2 10 20

order(x) # Indexes of sorted x elements

## [1] 4 2 1 3

names <- c("Maria", "Markus", "Lena")
sel <- c("Lena", "Maria")

match(sel, names) # Find 'sel' elements in 'names'

## [1] 3 1

x <- seq(1,100)
head(x,3)

## [1] 1 2 3

tail(x,5)

## [1]  96  97  98  99 100

A <- c(1, 2, 3, 4, 5, 6, 6)
B <- c(2, 4, 6, 8, 10)
union(A,B)

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

intersect(A,B)

## [1] 2 4 6

setdiff(A,B)

## [1] 1 3 5

helloWorld <- function() {

  outmessage <- "Hello World!\n"
  
  cat(outmessage)
  
}

helloWorld <- function() {

  outmessage <- "Hello World!\n"
  
  cat(outmessage)
  
}

helloWorld()

## Hello World!

helloWorld2 <- function(to = "World") {

  outmessage <- paste("Hello", to, "!\n", sep = " ")
  
  cat(outmessage)
  
}

helloWorld2 <- function(to="World") {

  outmessage <- paste("Hello", to, "!\n", sep=" ")
  
  cat(outmessage)
  
}

helloWorld2()

## Hello World !

helloWorld2("Francesca")

## Hello Francesca !

Let’s build a function that computes the body mass index (BMI) taking as arguments the weigth in kg and the height in meters.

BMI <- function (weight, height) {
  
  bmi <- weight/height^2
  
  return(bmi)
  
}

Note: the return function returns the final result/object produced by the function and stops any other evaluation. Any code written after the return statement is not evaluated.

We can calculate the BMI of an individual who weights 80 kg and is 1.90 m tall

BMI(80, 1.90)

## [1] 22.16066

To get meaningful results, we have to respect the argument order or specify their names

BMI(1.90, 80)

## [1] 0.000296875

BMI(height = 1.90, weight = 80)

## [1] 22.16066

R packages can be installed from CRAN or local archive files using the “Packages” panel of Rstudio

Packages > Install

By selecting one of the two following options

1. Package Archive File (.tgz; .tar.gz) > [Locate the archive file]
2. Install from Repository (CRAN) > [Specify the package name]

Or, for functions available on CRAN, by using directly the install.packages function (example installation of xlsx)

install.packages("xlsx")

Bioconductor is an open source software project for the analysis of genomic data with a repository containing >1500 R packages

Bioconductor pages contain different info about a package, including:

• Installation instructions
• Package manual and vignette
• Source code as archive files

R packages can be installed from Bioconductor with the following code (example installation of edgeR)

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("edgeR", version = "3.8")

GitHub is a provider of Internet hosting for software development and version control that can be used host R packages in public repositories.

R packages can be (tenatively) installed from GitHub using the install_github function from the devtools CRAN package (example installation of xCell)

library(devtools) 
install_github('dviraran/xCell')