Return the "sub-levels" within column B for each level of column A in R

Question

I have a dataset that looks like this:

  number  fruit status
1      1  apple   ripe
2      2  apple rotten
3      3 banana   ripe
4      4 banana rotten
5      5   pear   ripe
6      6   pear rotten
7      7  apple   ripe

 data.frame(number = 1:7,
       fruit = c(rep(c("apple","banana","pear"), each = 2),"apple"),
       status =c(rep(c("ripe", "rotten"),3),"ripe"))

I would like to loop over "fruit" and return the levels of "status" for each fruit. That is, get out something like this:

 $apple
 [1] ripe rotten

 $banana
 [2] ripe rotten

 $pear
 [3] ripe rotten

It doesn't have to be a list; I just need to know the levels within each "fruit" level. My data is more complicated than the example so assume I can't just remove the "number" column

I am trying to use apply functions or dplyr and I can't figure out how to get this.

Answer 1

1) tapply/unique Assuming only unique values of status are wanted this base R solution could be used:

with(DF, tapply(as.character(status), fruit, unique, simplify = FALSE))

giving:

$apple
[1] "ripe"   "rotten"

$banana
[1] "ripe"   "rotten"

$pear
[1] "ripe"   "rotten"

2) split If it were known that the sublevels of each level are already unique then this base R solution would be sufficient and gives the same result.

with(DF, split(as.character(status), fruit))

3) table Another form of output that might be useful is a table showing the number of occurrences of each sublevel within in each level. Again this uses only base R.

m <- table(DF[-1])
m

giving:

        status
fruit    ripe rotten
  apple     1      1
  banana    1      1
  pear      1      1

We can create a bipartite graph of this using the igraph package:

library(igraph)
g <- graph_from_incidence_matrix(m)
plot(g, layout = layout_as_bipartite)