CODEWITHSUNDEEP
rrecursiondataframenested-lists
Gabriele B
Gabriele B

Reputation: 2685

Convert recursive lists to dynamic data.frame in R

I need to build a function that traverses lists and sublists recursively and then convert to data.frame. Let's say I have a list like that

from <- list(id="12345678", name="Gabriele")
message <- "The quick fox"
comments <- list(list(name="Mickey", comment="Hello world!"), list(name="Donald", message="World...hello!"))

big.list <- list(from, message, comments)

I need to convert it in the form of a data.frame with this schema

from_id, from_name, message, comments_name, comments_message

(in other words, flattening the sublists)

The problem is that I don't know in advance which fields I have in the lists and which I've not (for example, some posts may miss the comments section altogether).

Upvotes: 2

Views: 517

Answers (1)

Jaehyeon Kim
Jaehyeon Kim

Reputation: 1417

Below is a quick trial. If big.list has names, they can be taken as column names.

The inner do.call() flattens the sub-list while the outer one converts into a data frame.

lst1 <- list(from = list(id="12345678", name="Gabriele"),message = "The quick fox",
             comments = list(list(name="Mickey", comment="Hello world!"),
                            list(name="Donald", message="World...hello!")))
lst2 <- list(from = list(id="12345678", name="Gabriele"),message = "The quick fox",
             comments = list(list(name="Mickey", comment="Hello world!")))
lst3 <- list(from = list(id="12345678", name="Gabriele"),message = "The quick fox")

df1 <- do.call(data.frame, do.call(c, lst1))
df2 <- do.call(data.frame, do.call(c, lst2))
df3 <- do.call(data.frame, do.call(c, lst3))

df1
#from.id from.name       message comments1.name comments1.comment comments2.name comments2.message
#1 12345678  Gabriele The quick fox         Mickey      Hello world!         Donald    World...hello!

df2
#from.id from.name       message comments.name comments.comment
#1 12345678  Gabriele The quick fox        Mickey     Hello world!

df3
#from.id from.name       message
#1 12345678  Gabriele The quick fox

@ Gabriele B

Below is clumsy but working solution. Others may post a better one.

from <- list(id="12345678", name="Gabriele")
message <- "The quick fox"
comments <- list(list(name="Mickey", comment="Hello world!"), list(name="Donald", message="World...hello!"))
big.list <- list(from = from, message = message, comments = comments)
big.list1 <- list(from = from, message = message)

join <- function(lst) {
  if(length(lst$comments) < 1) {
    bnd <- data.frame(id = unlist(lst$from)[1], name = unlist(lst$from)[2], lst$message)
    bnd$cmt.name <- "na"
    bnd$comment <- "na"
    bnd
  } else {
    bnd <- do.call(rbind, lapply(1:length(lst$comments), function(x) {
      id <- unlist(lst$from)[1]
      name <- unlist(lst$from)[2]
      data.frame(id = id, name = name, lst$message)
    }))
    bnd <- cbind(bnd, do.call(rbind, lst$comments))
    names(bnd) <- c("id", "name", "message", "cmt.name", "comment")
    bnd
  }  
}

join(big.list)
#id     name       message cmt.name        comment
#id  12345678 Gabriele The quick fox   Mickey   Hello world!
#id1 12345678 Gabriele The quick fox   Donald World...hello!

join(big.list1)
#id     name   lst.message cmt.name comment
#id 12345678 Gabriele The quick fox       na      na

Upvotes: 1

Related Questions