General R guidelines

Packages

To install a package available in CRAN

install.packages("ggplot2")

load the package using:

library(ggplot2) # Load
detach(package:ggplot2,unload=T) # Unload

Other packages like DESeq2 are only available in Bioconductor

you can install them like this:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("DESeq2")

If two package have the same function name use the format Package::Function, (e.g. above BiocManager::install(“DESeq2”))

If you want to know where your packages are, run this command:

.libPaths()

Help

You can see the help of a function using “?”

?write.table

Working directory

Always think where is your “working directory (wd)” in R

In an Rmd, the working directory is located in the same directory as the Rmd file

# Get the working directory
getwd()
## [1] "C:/Users/Maxime-Laptop/OneDrive - Université de Lausanne/Bureau/Introduction_R/Introduction_R/lectures/2.Intro_And_Variables"

You can change the working directory, but it will be only valid for a chunk

setwd(dir = "image/")
list.files(path=".")
## [1] "help.PNG"

if you want to change the wd for all the document:

# knitr::opts_knit$set(root.dir = '/tmp')

Comments

You can use comment in your code or Rmarkdown chunk using the # character

# This is a comment
1 + 1 
## [1] 2

Variable name

Some variable name should not be used:

# Legal variable names:
myvar <- "John"
my_var <- "John"
myVar <- "John"
MYVAR <- "John"
myvar2 <- "John"
.myvar <- "John"
# Illegal variable names:
2myvar <- "John"
my-var <- "John"
my var <- "John"
_my_var <- "John"
my_v@ar <- "John"
TRUE <- "John"

Variables in R

Numeric variables

Attribute value to a variable ( R is case sensitive !)

# Single variable
MyVariable<-15
# Vector 
ManyVariables<-c(1,2,3,4,5)

Integer

 
# Add L to force number to be integer
SomeNumbers<-c(1L,22L,185L)
SomeNumbers
## [1]   1  22 185
class(SomeNumbers)
## [1] "integer"

Double

 
SomeDoubleNumbers<-c(1.21, 2/3 , pi)
SomeDoubleNumbers
## [1] 1.2100000 0.6666667 3.1415927
class(SomeDoubleNumbers)
## [1] "numeric"

Boolean variables

Variables are only TRUE or FALSE ( Do not use quotes !)

MyBoolean<-c(TRUE,FALSE,TRUE,FALSE)
MyOtherBoolean<-c(TRUE,TRUE,FALSE,FALSE)

Some operation that return boolean:

AND

 
MyBoolean & MyOtherBoolean
## [1]  TRUE FALSE FALSE FALSE

OR

 
MyBoolean | MyOtherBoolean
## [1]  TRUE  TRUE  TRUE FALSE

NOT

 
!MyBoolean
## [1] FALSE  TRUE FALSE  TRUE

Compare values

 
SomeDoubleNumbers<-c(1.21, 2/3 , pi)
SomeNumbers<-c(1L,1L,1L)
SomeDoubleNumbers > SomeNumbers
## [1]  TRUE FALSE  TRUE

Characters & Factors

Often these two data types are confunded

Condition<-c("Ctr","Ctr","Treat","Treat")
Condition_AsFactor<-factor(Condition)

Character

 
# print character
Condition
## [1] "Ctr"   "Ctr"   "Treat" "Treat"
class(Condition)
## [1] "character"
class(Condition_AsFactor)
## [1] "factor"

Factor

 
# print factors
Condition_AsFactor
## [1] Ctr   Ctr   Treat Treat
## Levels: Ctr Treat
# At a lower level, factors are stored as integer
typeof(Condition_AsFactor)
## [1] "integer"

You can get the levels of a factor

levels(Condition_AsFactor)
## [1] "Ctr"   "Treat"

Vectors

You can only have one type of variable in a vector.

# Here numeric values are converted into characters
MyVectory<-c("A",1.2,2/4,2L)
MyVectory
## [1] "A"   "1.2" "0.5" "2"

To access a value in a vector use brackets “[]” (e.g. MyVectory[1])

You can access multiple elements

MyVectory[1:3]
## [1] "A"   "1.2" "0.5"
# To access the last element of a vector: 
# MyVectory[length(MyVectory)] or tail(MyVectory,1)

Vectors can be named (often required in packages)

MyGenes<-c("Arntl"=1.5,"Per2"=1.8,"Cdk4"=8.5)
MyGenes
## Arntl  Per2  Cdk4 
##   1.5   1.8   8.5

You can change the names of the vectors

names(MyGenes)
## [1] "Arntl" "Per2"  "Cdk4"
# Change name 2
names(MyGenes)[2]<-"Cry1"
MyGenes
## Arntl  Cry1  Cdk4 
##   1.5   1.8   8.5

Lists

If you want multiple type of variable, you can use lists.

You don’t need same length for each element in the list

MyList<-list("SomeNumber"=c(1.5,2.3,4.5),
             "SomeName"=c("Gene1","Gene2","Gene3","Gene4","Gene5"),
             "SomeFactors"=factor(c("C1","C1","C2","C4")))

You can see the names of elements in the list

names(MyList)
## [1] "SomeNumber"  "SomeName"    "SomeFactors"

You can access the “Numbers” of the list using brackets

But, it will return a list of 1 element 

MyList["SomeNumber"]
## $SomeNumber
## [1] 1.5 2.3 4.5

To access the vector use double brackets or $

MyList$SomeNumber
## [1] 1.5 2.3 4.5
MyList[["SomeNumber"]]
## [1] 1.5 2.3 4.5

Matrix

Matrix contains only one type of variable

# Completion is done by column
MyMatrix<-matrix(c(1,2,3,4,5,6),ncol=2)
MyMatrix
##      [,1] [,2]
## [1,]    1    4
## [2,]    2    5
## [3,]    3    6

You can change rownames and colnames

colnames(MyMatrix)<-c("A","B")
rownames(MyMatrix)<-c("C","D","E")

You can acces a specific column or row

If only 1 row/column is requested, R return a vector. Think to use ,drop=F !

MyMatrix[1,] # First row
## A B 
## 1 4
MyMatrix[,1] # First column
## C D E 
## 1 2 3
  • MyMatrix[,1] will return a vector
  • MyMatrix[,1,drop=F] will return a 2-dimension item
  • MyMatrix[,1:2] will return a 2-dimension item

Data.frame

Similar to matrix, but can contain multiple type of data

MyDataFrame<-data.frame("Read_counts"=rpois(4,lambda = 4),
                        "Genes"=c("Cry1","Per2","Arntl","Cdk4"))
MyDataFrame
##   Read_counts Genes
## 1           1  Cry1
## 2           6  Per2
## 3           7 Arntl
## 4           3  Cdk4

Again, be carefull to variable type conversion when selecting a single item !

Selecting 1 row kept the data.frame format

class(MyDataFrame[1,])
## [1] "data.frame"

but not when selecting 1 column

# Same if you use class( MyDataFrame$Read_counts )
class(MyDataFrame[,1])
## [1] "integer"

cbind and rbind

You may want to add new data to your data.frame or matrix, you can use rbind to add new line or cbind to add new row.

MyDataFrame
##   Read_counts Genes
## 1           1  Cry1
## 2           6  Per2
## 3           7 Arntl
## 4           3  Cdk4
MyDataFrame<-cbind(MyDataFrame,PassQC=c(T,F,F,T))
MyDataFrame
##   Read_counts Genes PassQC
## 1           1  Cry1   TRUE
## 2           6  Per2  FALSE
## 3           7 Arntl  FALSE
## 4           3  Cdk4   TRUE
MyDataFrame<-rbind(MyDataFrame,list(3,"P53",T))
MyDataFrame
##   Read_counts Genes PassQC
## 1           1  Cry1   TRUE
## 2           6  Per2  FALSE
## 3           7 Arntl  FALSE
## 4           3  Cdk4   TRUE
## 5           3   P53   TRUE

Indexing

To acces certain information, you can in use logical indexing. e.g. filter low count reads:

MyDataFrame
##   Read_counts Genes PassQC
## 1           1  Cry1   TRUE
## 2           6  Per2  FALSE
## 3           7 Arntl  FALSE
## 4           3  Cdk4   TRUE
## 5           3   P53   TRUE

Here we keep only gene with a minimum of 3 counts

MyDataFrame[MyDataFrame$Read_counts > 3,]
##   Read_counts Genes PassQC
## 2           6  Per2  FALSE
## 3           7 Arntl  FALSE

You can see the logical index that was given

MyDataFrame$Read_counts > 3
## [1] FALSE  TRUE  TRUE FALSE FALSE

This is equivalient to give:

MyDataFrame[c(FALSE,TRUE,TRUE,FALSE),]
##   Read_counts Genes PassQC
## 2           6  Per2  FALSE
## 3           7 Arntl  FALSE

Some operator

  • “>” more than
  • “<” less than
  • “==” equal (but see which)
  • “!=” not equal

Which functions

The standard operator can be sometime problematic with NA

MyGenes<-c("Arntl"=1.5,"Per2"=1.8,"Cdk4"=8.5,"Cry1"=NA)
MyGenes
## Arntl  Per2  Cdk4  Cry1 
##   1.5   1.8   8.5    NA

it will return NA value

MyGenes[MyGenes>2]
## Cdk4 <NA> 
##  8.5   NA

you can easily use which

MyGenes[which(MyGenes>2)]
## Cdk4 
##  8.5

it return the an index which are TRUE

which(MyGenes>2)
## Cdk4 
##    3

you can also use it for min or max value

  • which.min(MyGenes)
  • which.max(MyGenes)

The %in% operator

You can return index for item that match multiple requirement

MyGenes<-c("Arntl"=1.5,"Per2"=1.8,"Cdk4"=8.5,"Cry1"=NA)
MyGenesName<-names(MyGenes)

Find the names equal to Arntl and Per2

idx<-MyGenesName %in% c("Arntl","Per2")
MyGenesName[idx]
## [1] "Arntl" "Per2"

or genes with specific values

idx<-MyGenes %in% c(1.5,8.5)
MyGenesName[idx]
## [1] "Arntl" "Cdk4"

Remove item in list

You can remove some item by using the ‘-’ sign

MyGenes<-c("Arntl"=1.5,"Per2"=1.8,"Cdk4"=8.5,"Cry1"=NA)

Remove genes with value < 1.8

MyGenes[-which(MyGenes<1.8)]
## Per2 Cdk4 Cry1 
##  1.8  8.5   NA

also remove NA

# less than 1.8 OR value equal NA
MyGenes[-which(MyGenes<1.8 | is.na(MyGenes))]
## Per2 Cdk4 
##  1.8  8.5

Transform data

as.numeric(c("1","2","3"))
## [1] 1 2 3
as.character(22:25)
## [1] "22" "23" "24" "25"

Exercice 2