Show Menu
Cheatography
  1. Home >
  2. Cheat Sheets >

R Base Cheat Sheet (DRAFT) by

Beginners guide for R

This is a draft cheat sheet. It is a work in progress and is not finished yet.

Shor­tcuts

cmd + enter
Runs the line you are in and goes to the next line
 
alt + enter
Runs the line you are in

Getting help

?
(Looks in your library)
 
??
(Looks in every library)
 
help.s­ear­ch(­"­_")
looks for a word or phrase
 
help(p­ack­age­="_")
find help for a package
Getting help is very useful when you are defining functions or other things, it provides inform­ation about all the different conditions possible.

Working direct­ory

getwd()
Tells you your working directory
 
setwd(­'C:­//f­ile­/path')
How to code your desired wd.
You can select it manually in RStudio by clicking in the gear.

It has to be set to open correctly a .csv, if the file you desire to open is not in the same folder where you are saving your code you are going to need to define the path for opening without any problems.

Working with .csv

To open a .csv:
Whenever you open a csv, you have to save them with some name in order to be able to access to it and have it in your enviro­nment
 
read.csv(path/name of the csv, ... )
When opening a csv, look at the help of the function in order to see all the needs for each case.
You can open manually a csv by clicking the import button in the enviro­nment. For this the 
readr
package is needed.
Result: 
ds<­-re­ad.c­sv()
its always good to open the imported dataset to see if everything is correct. 
View()
To save a .csv:
 
write.c­sv(df, "­__",...)
When saving a .csv look into the help to see all the possible variables that can be rearrange.

Put things together

paste(­"­"­,"", sep=" ")
Pastes two things together
 
paste0()
Pastes things without separator

Creating functions

myfunction<- function (x){
 code 
 paste() 
}
Difference between paste, print and return

What happens inside the function stays in the function

Dates

as.Date( __, format = "­__")


In the first blank you have to put your dates, in the format how R needs to read it.

RStudio unders­tands that dates can exist and it knows how they work. when a variable is a date is going to show in the form :
year-m­ont­h-day
we can change this when we transform it into a character. Also, and because R unders­tands dates as said before, when you are defining it from a string you have to explain how to read it.

To get current date and time:

Sys.Date()
You can ask for a sequence of dates if you use 
seq(as.Da­te(­),a­s.D­ate­(),by= " ")
and select if you want it to be daily, weekly, monthly etc by saying so in the by= "­day­"

Lists

list(_ , _ , _ )
To create a colum
 
list[[1]]
To access to the first element of the list
 
list[[­1]][1]
To access to the first element of the first element of the list
To combine lists 
<-c­(li­st1­,list2)

Tables of propor­tions

prop.t­able(t)
It sums up to 1
 
prop.t­able(t, 2)
Column­-wise
 
prop.t­able(t, 1)
Row-wise

Working with NA

When working with NA's a lot of different operations don't work, because of that you have to ask for different things to obtain results
 
any.NA()
It returns TRUE if one of the elements is NA
 

Types of data

Logical
 
TRUE, FALSE
Boolean Values (T/F).
Numeric
 
2, 5, 7
Interger or floating point numbers.
Character
 
"­hel­lo", "­bye­"
Character strings.
Factor
 
"­mal­e", "­fem­ale­"
Character strings with different levels.
 
different levels
 
levels(_)
For assessing the levels
  
NA
Missing values
For changing for one type to other you have to use the function 
as._()
and it will transform in the type you decide. You can also transform dates.

Cond­iti­ons

!
Not
 
a == b
Are equal
 
a != b
Are not equal
 
a > b
Great than
 
a < b
Less than
 
a >= b
Greater or equal
 
a <= b
Less or equal
 
is.na(a)
Is missing
 
is.null(a)
Is null
 
&
And
 
|
Or

Operations with characters

substr­ing­("__­", first= #, last= #)
Returns the characters inside the string within those positions
 
nchar()
Counts the number of characters (including symbols and spaces)

Reshaping DataFrames

melt
 
(df, id.var­s="", variab­le.n­am­e="")
Transforms to long
 
dcast
 
(df, id~mea­sure)
Transform to wide
To ask for help look for reshape2

Look session 5

Table

prop.t­able(t)
Returns a table with the propor­tions of all
 
prop.t­abl­e(t,2)
Propor­tions by colums
 
prop.t­abl­e(t,1)
Propor­tions by rows

To round numbers:

round (x, #)
It will round the x you ask with the decimals you ask (#)
 
ceiling(#)
It will round to the next number
 
floor(#)
It will round in the number you have

Condit­ionals

if (<c­ond­iti­on>) { 
  <co­de> 
}
Just one condition
 
if (<c­ond­iti­on>) {  
 <code>
 } else {
   <co­de> 
}
One condition, if not the rest without condition
 
if (<c­ond­iti­on>) { 
  <co­de> 
} else if (<c­ond­iti­on>) {
   <co­de> 
} else {
   <co­de> 
}
More than one condition
You have to be careful when you place your condit­ions, the one that conditions the most have to be the first and so on

Creating Data Frames

df<- data.frame("name1"=c(values),
 "­nam­e2"=­c(v­alues), etc)
You have to define the column names and give it the values that you want. The values can be a vector, a list or other things.
For getting things from data sets: 
df[[_]]
You can put the number of the row but is better putting the name.
Is very useful to use the command $ to access a data frame 
df$col­umname
For adding things:
 
rbind (df1,df2)
Fails: row numbers don't match
 
cbind ( df1,df2)
Fails: column names don't match
 
aggreg­ate()
session 6 To do aggreg­ations and get the result formatted as a data.frame
 

Apply

lapply( __, function)
apply a function over a list or a vector
 
sapply(__, function)
same as lapply but with simplified results (better)
 
tapply(__, groupi­ng,­fun­ction)
Apply a function over a ragged array
Session 6

Math

sqrt()
Square root of the number
 
log()
Logarithm of the number
 
abs()
Absolute value
 
+
Sum
 
-
Substr­action
 
*
Multip­lic­ation
 
/
Division
 
^
Expone­ntial
 
%%
Module operator
 
union()
Union
 
inters­ect()
Inters­ection
 
setdiff()
Difference
 
%in%
Membership
 
pct=TRUE
Percentage (*100)
 
mean()
Mean
 
median()
Median
 
sd()
Standar deviation
 
quantile()
Quantiles
 
quanti­le(­df$col, 
seq(0, 1, 0.1))
Percen­tiles
 
cor()
Correl­ation between variables

Functions

summar­y(df)
Gives you info about all the columns, (min, max, median, mean, 1n3Q)
 
head(df)
Gives you the first 6 lines by default, you can change it.
 
tail(df)
Gives you the last 6 lines by default, you can change it.
 
dim(df)
Gives you the dimensions of your df
 
str(df)
Gives you like a list with the variables
 
barplot(t)
Creates a barplot, not the way we are going to do them
 
length ( )
Returns the length
Logical operators:
 
all( )
TRUE if all elements are true
 
any( )
TRUE if any of the elements is true
To repeat things
 
rep()
 
rep( _ , times=#)
You repeat that same thing the numbers you asked
 
rep( _ , each= #)
You repeat each thing # times
 
sum()
Sum of vector elements
 
seq()
or 
_:_
Generates a sequence
 
cumsum()
Cumulative sum in each position
 
diff()
Like cumsum but substr­acting
 
nchar()
Counts the number of characters in each position
 
grep("_­", vector­,ig­nor­e.c­ase­=TRUE)
Pattern Matching and Replac­ement -> returns position
 
grepl(­"­_",v­ect­or,­ign­ore.ca­se=­TRUE)
Pattern Matching and Replac­eme­nt-> returns logical vector
 
gsub("",­"­",_, ignore.ca­se=­FALSE)
For replac­ement, first what you want to put out, the what you want to put, and then where.

Data filtering and reordering

You can use logical condit­ions, they can be used in two forms:
- By creating a logical vector and applying it
 
logica­l_v­ector <- c(TRUE, FALSE, TRUE, FALSE) 
products_stock[logical_vector]
-By just applying the logical condition
 
vector­[c(­TRUE, FALSE, TRUE, TRUE, FALSE)]
- With the function 
which()
returns the positions
You can put condit­ions:
- Either inside the line: 
df[df$col1 < 25, ]
- Or with function subset()
 
subset( df, column name with the condition)
With subset you can put more than one condition with the command 
&
Reorde­ring:
- 
sort ( _,  decrea­sin­g=F­ALSE)
by default decreasing is false and you can omit it. It rearrange the vector
- 
 order ( _, decrea­sin­g=F­ALSE)
by default decreasing is false and you can omit it. It just gives the positions where they should be