CODEWITHSUNDEEP
rlistjsonlite
User9123
User9123

Reputation: 13

R: Combine lists of same length into list of tupels

I'm currently struggling with the following problem in R: Given two lists a=(a1,...,an) and b=(b1,...,bn), I want to get a list of lists like ((a1,b1),...,(an,bn)). Background info: I need to convert a plain data table/frame into a nested json (e.g. using jsonlite). As an code example consider a data table with ids, latitudes and longitudes, the latter two to be summarized in a list called location:

library(data.table)
n<-5
data<-data.table(id=1:n,lon=1:n,lat=1:n)

Here, we could use lapply to get the required result:

data$location<-lapply(1:nrow(data),function(x) list(data[x,c("lat","lon"),with=F]) )

Alternatively split (faster for small, slower for bigger data sets):

data$location<-list(split(data[,c("lat","lon"),with=F],1:nrow(data)))

Both work fine on small scale, but for n>>10^5 it takes ages on my machine. Do you have any solution in mind for faster computation?

Upvotes: 1

Views: 613

Answers (2)

akrun
akrun

Reputation: 887541

We can place the Subset of Data.table in a list after grouping by 'id' in 'data.table' and extract the default new column 'V1'

data[, location := list(list(.SD)), id]

We can extract the list column 

data$location
#[[1]]
#   lon lat
#1:   1   1

#[[2]]
#   lon lat
#1:   2   2

#[[3]]
#   lon lat
#1:   3   3

#[[4]]
#   lon lat
#1:   4   4

#[[5]]
#   lon lat
#1:   5   5

Upvotes: 1

akuiper
akuiper

Reputation: 215057

You can use data.table::transpose as another option:

data[, location := transpose(.(lon, lat))]

data
#   id lon lat location
#1:  1   1   1      1,1
#2:  2   2   2      2,2
#3:  3   3   3      3,3
#4:  4   4   4      4,4
#5:  5   5   5      5,5

data$location        # this drops the column names, you can refer to the elements by index
#[[1]]
#[1] 1 1

#[[2]]
#[1] 2 2

#[[3]]
#[1] 3 3

#[[4]]
#[1] 4 4

#[[5]]
#[1] 5 5

Upvotes: 0

Related Questions