Reputation: 1547
I would like to get all the common elements within all the possible combinations of n variables.
Let's say we have 3 variables: a, b and c, with 4, 2 and 3 elements each. I will like to create a data frame with all possible combinations in a column (a,b,c,ab,ac,bc and abc) and the common elements in a second column. Example:
input:
a<-c("x1","x2","x3","x4")
b<-c("x2","x5")
c<-c("x3","x4","x6")
I will like an output that looks something like:
name elements
1 a "x1","x2","x3","x4"
2 b "x2","x3","x5"
3 c "x3","x4","x6"
4 ab "x1","x2","x3","x4","x5"
5 ac "x1","x2","x3","x4","x6"
6 bc "x2","x3","x4","x5","x6"
7 abc "x1","x2","x3","x4","x5","x6"
So far I tried this:
all1<-c("a","b","c")
all2<-list(a,b,c)
all3<-c(a,b,c)
sapply ( 1:3, function(x) combn(all1,x))
sapply ( 1:3, function(x) combn(all2,x))
sapply ( 1:3, function(x) combn(all3,x))
Upvotes: 2
Views: 104
Reputation: 9705
Lots of ways, here's one:
grid <- expand.grid(a=c(T,F), b=c(T,F), c=c(T,F))
grid$results <- mapply(function(x,y,z) {
unique(c(if(x) a,if(y) b,if(z) c))
},grid$a,grid$b,grid$c)
Output:
a b c results
1 TRUE TRUE TRUE x1, x2, x3, x4, x5, x6
2 FALSE TRUE TRUE x2, x5, x3, x4, x6
3 TRUE FALSE TRUE x1, x2, x3, x4, x6
4 FALSE FALSE TRUE x3, x4, x6
5 TRUE TRUE FALSE x1, x2, x3, x4, x5
6 FALSE TRUE FALSE x2, x5
7 TRUE FALSE FALSE x1, x2, x3, x4
8 FALSE FALSE FALSE NULL
Upvotes: 4