I have a large data.frame like this:
+--------+---------+
| A | B |
+--------+---------+
| USA | Chicago |
+--------+---------+
| USA | Chicago |
+--------+---------+
| France | Paris |
+--------+---------+
| Italy | Rome |
+--------+---------+
| France | Nice |
+--------+---------+
| Italy | Venice |
+--------+---------+
ie
AB <- structure(list(A = c("USA", "France", "Italy", "France", "Italy",
"USA"), B = c("Chicago", "Paris", "Rome", "Nice", "Venice", "Chicago"
)), row.names = c(NA, -6L), class = "data.frame")
and I would like to create a list like this:
list(USA = list("Chicago"), France = list("Paris", "Nice"), Italy = list(
"Rome", "Venice"))
Here's what I'm doing now.
unique.As <- unique(AB$A)
ABL <- lapply(unique.As, function(current.A) {
return(unique(AB$B[AB$A == current.A]))
})
names(ABL) <- unique.As
I previously wrote that listifying a data.frame with 65k rows took ~ 10 minutes. I realized today that almost all of that time was from another step in the lapply loop that I didn't show above.
akrun's solution below is still faster and more elegant!
split
should be faster
lst1 <- split(as.list(AB$B), AB$A)
If the intention is to have both 'key', 'value' unique,
lst1 <- with(unique(AB), split(as.list(B), A))
Or
with(AB[!duplicated(AB), ], split(as.list(B), A))