第1组的唯一值，然后是第1组和第2组，依此类推

Question

I have a dataframe with 5 different groups :

   id group
1  L1     1
2  L2     1
3  L1     2
4  L3     2
5  L4     2
6  L3     3
7  L5     3
8  L6     3
9  L1     4
10 L4     4
11 L2     5

I would like to know if it's possible to get the unique id from the 1st group, the 1st and the 2nd, the 1st, 2nd and 3rd and so on without for looping. I'm searching a way with dplyr or data.table package.

Expected results :

    group      id
1   1          c("L1", "L2")
2   1,2        c("L1", "L2", "L3", "L4")
3   1,2,3      c("L1", "L2", "L3", "L4", "L5")
4   1,2,3,4    c("L1", "L2", "L3", "L4", "L5")
5   1,2,3,4,5  c("L1", "L2", "L3", "L4", "L5")  

Data :

structure(list(id = c("L1", "L2", "L1", "L3", "L4", "L3", "L5", 
"L6", "L1", "L4", "L2"), group = structure(c(1L, 1L, 2L, 2L, 
2L, 3L, 3L, 3L, 4L, 4L, 5L), .Label = c("1", "2", "3", "4", "5"
), class = "factor")), .Names = c("id", "group"), row.names = c(NA, 
-11L), class = "data.frame")

Answer 1

With base R, you can do:

# create the "growing" sets of groups
combi_groups <- lapply(seq_along(unique(df$group)), function(i) unique(df$group)[1:i])

# get the unique ID for each set of groups
uniq_ID <- setNames(lapply(combi_groups, function(x) unique(df$id[df$group %in% x])), 
                    sapply(combi_groups, paste, collapse=","))

# $`1`
# [1] "L1" "L2"

# $`1,2`
# [1] "L1" "L2" "L3" "L4"

# $`1,2,3`
# [1] "L1" "L2" "L3" "L4" "L5" "L6"

# $`1,2,3,4`
# [1] "L1" "L2" "L3" "L4" "L5" "L6"

# $`1,2,3,4,5`
# [1] "L1" "L2" "L3" "L4" "L5" "L6" 

If you want to format as in your expected output:

data.frame(group=sapply(combi_groups, paste, collapse=", "), id=sapply(uniq_ID, function(x) paste0("c(", paste0("\"", x, "\"", collapse=", "), ")")))
#          group                                    id
#1             1                         c("L1", "L2")
#2          1, 2             c("L1", "L2", "L3", "L4")
#3       1, 2, 3 c("L1", "L2", "L3", "L4", "L5", "L6")
#4    1, 2, 3, 4 c("L1", "L2", "L3", "L4", "L5", "L6")
#5 1, 2, 3, 4, 5 c("L1", "L2", "L3", "L4", "L5", "L6")

Another possibility of formatting:

data.frame(group=rep(names(uniq_ID), sapply(uniq_ID, length)), id=unlist(uniq_ID))

Or, if you want to have uniq_ID in a column:

library(data.table)
data.table(group=sapply(combi_groups, paste, collapse=", "), id=uniq_ID)
#           group                id
#1:             1             L1,L2
#2:          1, 2       L1,L2,L3,L4
#3:       1, 2, 3 L1,L2,L3,L4,L5,L6
#4:    1, 2, 3, 4 L1,L2,L3,L4,L5,L6
#5: 1, 2, 3, 4, 5 L1,L2,L3,L4,L5,L6

data.table(group=sapply(combi_groups, paste, collapse=", "), id=uniq_ID)[2, id]
[[1]]
[1] "L1" "L2" "L3" "L4"

Answer 2

In similar vein as the answer of @Cath, but using Reduce(..., accumulate = TRUE) to create the expanding window of groups. Then loop over the set of groups with lapply to get the unique id's for each window:

grp <- Reduce(c, unique(d$group), accumulate = TRUE)

lapply(grp, function(x) unique(d$id[d$group %in% x]))
# [[1]]
# [1] "L1" "L2"
# 
# [[2]]
# [1] "L1" "L2" "L3" "L4"
# 
# [[3]]
# [1] "L1" "L2" "L3" "L4" "L5" "L6"
# 
# [[4]]
# [1] "L1" "L2" "L3" "L4" "L5" "L6"
# 
# [[5]]
# [1] "L1" "L2" "L3" "L4" "L5" "L6"

---

For naming and prettification, please refer to the nice answer by @Cath.

Answer 3

Another method is to use split and Reduce to feed the groups to union with accumulate=TRUE:

Reduce(union, split(df$id, df$group), accumulate=TRUE)
[[1]]
[1] "L1" "L2"

[[2]]
[1] "L1" "L2" "L3" "L4"

[[3]]
[1] "L1" "L2" "L3" "L4" "L5" "L6"

[[4]]
[1] "L1" "L2" "L3" "L4" "L5" "L6"

[[5]]
[1] "L1" "L2" "L3" "L4" "L5" "L6"