> Doing j by Group > Advanced Data Table Operations

R For Data Science > DT[,.(V4.Sum=sum(V4)),by=V1] #Calculate sum of V4 for every group in V1 Sum

V1 V4.

>
>
DT[.N-1] #Return the penultimate row of the DT

DT[,.N] #Return the number of rows

data.table Cheat Sheet

1: 1 36
> DT[,.(V2,V3)] #Return V2 and V3 as a data.table

2: 2 42
> DT[,list(V2,V3)] #Return V2 and V3 as a data.table

> DT[,.(V4.Sum=sum(V4)), by=.(V1,V2)] #Calculate sum of V4 for every group in V1 and V2

#Return the result of j, grouped by all possible combinations of groups specified in by

> DT[,.(V4.Sum=sum(V4)), by=sign(V1-1)] #Calculate sum of V4 for every group in sign(V1-1)

> DT[,mean(V3),by=.(V1,V2)]

sign V4.Sum
V1 V2 V1

Learn data.table online at www.DataCamp.com 1: 0

2: 1
36

42

1: 1
2: 1
A
B
0.4053

0.4053

#The same as the above, with new name for the variable you’re grouping by
3: 1 C 0.4053

> DT[,.(V4.Sum=sum(V4)), by=.(V1.01=sign(V1-1))]

4: 2 A -0.6443

#Calculate sum of V4 for every group in V1 after subsetting on the first 5 rows
5: 2 B -0.6443

> DT[1:5,.(V4.Sum=sum(V4)), by=V1]

6: 2 C -0.6443

data.table
> DT[,.N,by=V1] #Count number of rows for every group in V1

General form: DT[i, j, by] “Take DT, subset rows using i, then calculate j grouped by by” .SD & .SDcols
data.table is an R package that provides a high-performance
> DT[,print(.SD),by=V2] #Look at what .SD contains

version of base R’s data.frame with syntax and feature enhancements

for ease of use, convenience and programming speed. > Adding/Updating Columns By Reference in j Using := >
>
>
DT[,.SD[c(1,.N)],by=V2] #Select the first and last row grouped by V2

DT[,lapply(.SD,sum),by=V2] #Calculate sum of columns in .SD grouped by V2

DT[,lapply(.SD,sum),by=V2, #Calculate sum of V3 and V4 in .SD grouped by V2

.SDcols=c("V3","V4")]

Load the package: > DT[,V1:=round(exp(V1),2)] #V1 is updated by what is after :=

V2 V3 V4

> DT Return the result by calling DT

1: A -0.478 22

> library(data.table) V1 V2 V3 V4
2: B -0.478 26

1: 2.72 A -0.1107 1
3: C -0.478 30

2: 7.39 B -0.1427 2
> DT[,lapply(.SD,sum),by=V2, #Calculate sum of V3 and V4 in .SD grouped by V2

3: 2.72 C -1.8893 3
.SDcols=paste0("V",3:4)]
> Creating A data.table 4: 7.39
...

A -0.3571 4

> DT[,c("V1","V2"):=list(round(exp(V1),2), #Columns V1 & V2 are updated by what is after :=

> set.seed(45L) #Create a data.table and call it DT

> DT <- data.table(V1=c(1L,2L),

LETTERS[4:6])]

#Alternative to the above one. With [], you print the result to the screen

> DT[,':='(V1=round(exp(V1),2),

> Indexing And Keys

V2=LETTERS[1:3],
V2=LETTERS[4:6])][]

V3=round(rnorm(4),4),
V1 V2 V3 V4
> setkey(DT,V2) #A key is set on V2; output is returned invisibly

V4=1:12) 1: 15.18 D -0.1107 1

> DT["A"] Return all rows where the key column (set to V2) has the value A

2: 1619.71 E -0.1427 2
V1 V2 V3 V4

3: 15.18 F -1.8893 3
1: 1 A -0.2392 1

2: 2 A -1.6148 4

> Subsetting Rows Using i

4: 1619.71 D -0.3571 4

> DT[,V1:=NULL] Remove V1

3: 1 A 1.0498 7

> DT[,c("V1","V2"):=NULL] #Remove columns V1 and V2

4: 2 A 0.3262 10

> Cols.chosen=c("A","B")
> DT[c("A","C")] #Return all rows where the key column (V2) has value A or C

> DT[3:5,] #Select 3rd to 5th row

> DT[,Cols.Chosen:=NULL] #Delete the column with column name Cols.chosen
> DT["A",mult="first"] #Return first row of all rows that match value A in key column V2

> DT[3:5] #Select 3rd to 5th row

> DT[,(Cols.Chosen):=NULL] #Delete the columns specified in the variable Cols.chosen > DT["A",mult="last"] #Return last row of all rows that match value A in key column V2

> DT[V2=="A"] #Select all rows that have value A in column V2

> DT[c("A","D")] #Return all rows where key column V2 has value A or D

> DT[V2 %in% c("A","C")] #Select all rows that have value A or C in column V2 V1 V2 V3 V4

1: 1 A -0.2392 1

> set()-Family 2: 2
3: 1
A
A
-1.6148
1.0498
4

> Manipulating on Columns in j 4: 2

5: NA D
A 0.3262
NA
10

NA

set() > DT[c("A","D"),nomatch=0] #Return all rows where key column V2 has value A or D

V1 V2 V3 V4

> DT[,V2] Return V2 as a vector

1: 1 A -0.2392 1

[1] “A” “B” “C” “A” “B” “C” ...

Syntax: for (i in from:to) set(DT, row, column, new value)
2: 2 A -1.6148 4

> DT[,.(V2,V3)] #Return V2 and V3 as a data.table

> rows <- list(3:4,5:6)

3: 1 A 1.0498 7

> DT[,sum(V1)] #Return the sum of all elements of V1 in a vector

> cols <- 1:2

4: 2 A 0.3262 10

[1] 18

#Sequence along the values of rows, and for the values of cols,
#Return total sum of V4, for rows of key column V2 that have values A or C

#Return the sum of all elements of V1 and the std. dev. of V3 in a data.table

set the values of those elements equal to NA (invisible)

> DT[c("A","C"),sum(V4)]

> DT[,.(sum(V1),sd(V3))]

> for(i in seq_along(rows))

#Return sum of column V4 for rows of V2 that have value A, sum(V4),

V1 V2

{set(DT,
and anohter sum for rows of V2 that have value C

1: 18 0.4546055

i=rows[[i]],
> DT[c("A","C"), by=.EACHI]

> DT[,.(Aggregate=sum(V1), #The same as the above, with new names

j=cols[i],
V2 V1

Sd.V3=sd(V3))]

value=NA)} 1: A 22

Aggregate Sd.V3

2: C 30

1: 18 0.4546055

> setkey(DT,V1,V2) #Sort by V1 and then by V2 within each group of V1 (invisible)

#Select column V2 and compute std. dev. of V3, which returns a single value & gets recycled

> DT[,.(V1,Sd.V3=sd(V3))]
setnames() #Select rows that have value 2 for the first key (V1) &

the value C for the second key (V2)

> DT[,.(print(V2), #Print column V2 and plot V3

> DT[.(2,"C")]

plot(V3),
Syntax: setnames(DT,"old","new")[]
V1 V2 V3 V4

NULL)] > setnames(DT,"V2","Rating") #Set name of V2 to Rating (invisible)

1: 2 C 0.3262 6

> setnames(DT, #Change 2 column names (invisible)

2: 2 C -1.6148 12

c("V2","V3"),
Select rows that have value 2 for the first key (V1) &

> Chaining
c("V2.rating","V3.DC")) within those rows the value A or C for the second key (V2)

> DT[.(2,c("A","C"))]

V1 V2 V3 V4

setcolorder() 1: 2 A -1.6148 4

> DT <- DT[,.(V4.Sum=sum(V4)), by=V1] #Calculate sum of V4, grouped by V1

2: 2 A 0.3262 10

V1 V4.Sum
3: 2 C 0.3262 6

Syntax: setcolorder(DT,"neworder")
1: 1 36
4: 2 C -1.6148 12
2: 2 42
> setcolorder(DT, #Change column ordering to contents of the specified vector (invisible)

> DT[V4.Sum>40] #Select that group of which the sum is >40

c("V2","V1","V4","V3"))
> DT[,.(V4.Sum=sum(V4)), #Select that group of which the sum is >40 (chaining)

by=V1][V4.Sum>40]

V1 V4.Sum

1: 2 42

> DT[,.(V4.Sum=sum(V4)), by=V1][order(-V1)] Calculate sum of V4, grouped by ordered on V1

V1 V4.Sum

1: 2
2: 1
42

36
Learn Data Skills Online at www.DataCamp.com