Normalize data frame with list column

Question

I have a data frame that maps some ids to a list of versions:

id versions
 1  1, 2, 4
 2        1
 3     3, 4

It can be created with the following code:

df <- data.frame(id=c(1, 2, 3), 
  versions=c("1 2 4", "1", "3 4"), 
  stringsAsFactors=F)
df$versions <- strsplit(df$versions, " ")

Notice that each element of the versions column is a list.

How to normalize this data frame? I need to get a data frame like this:

hmm, lots of ways of asking the same question – rodrigorgs Mar 02 '14 at 14:52 — rodrigorgs, Mar 02 '14 at 14:52

score 4 · Accepted Answer · answered Mar 02 '14 at 15:52

4

stack would be perfect for this:

stack(setNames(df$versions, df$id))
#   values ind
# 1      1   1
# 2      2   1
# 3      4   1
# 4      1   2
# 5      3   3
# 6      4   3

answered Mar 02 '14 at 15:52

A5C1D2H2I1M1N2O1R2T1

190,393
28
405
485

Nice! I'd just add `colnames(x) <- c("version", "id"); x <- x[, c("id", "version")]` to format the data frame `x` according to the question (assuming `x` is the data frame returned by `stack()`). – rodrigorgs Mar 03 '14 at 10:52

score 1 · Answer 2 · edited May 23 '17 at 11:45

I adapted and simplified the solution from another SO question for future reference:

data.frame(id = rep(df$id, sapply(df$versions, length)),
      version = unlist(df$versions))

The new id column is computed by repeating each id according to the number of versions it has (i.e., the length of the list versions). The new version column is computed using unlist, that just returns a vector by concatenating all elements in the list.

Normalize data frame with list column

2 Answers2