do.call from apply function

[masarimk]

Dear All,

I would like to cluster each column of below matrix. Then I want to extract cluster vectors as column matrix. Thank you

 df = data.frame(id=c(3,5,1,6,8,2,3,5), val=c(14,17,11,19,12,13,17,16))
ss=apply(df, 2, function(x) kmeans(x,3))
ff=do.call(c,ss) 

$val.cluster
[1] 2 3 2 1 2 2 3 3
 $id.cluster
[1] 2 3 1 3 3 1 2 3

But the resulst should look like
$val.cluster   $id.cluster
2                      2
3                      3
2                      1
1                      3
2                      3
2                      1
3                      2
3                      3 

[Dason]

It sounds more like you want something like this

do.call(cbind, sapply(ss, "[", 1))

The sapply(ss, "[", 1) grabs the first element from each part of the list called ss. Then we want to bind those together as columns so we use cbind. But we want to bind the elements of a list together so we need to use do.call so those get passed as the parameters to cbind.

[masarimk]

Thank you very much. I really appreciate that. One more question:
When I run below code , it gives me X1 and X2 not the variable names,Is there a way that I can put the names of the variables for the center?

df = data.frame(id=c(3,5,1,6,8,2,3,5), val=c(14,17,11,19,12,13,17,16))
ss=apply(df, 2, function(x) kmeans(x,3))
data.frame(do.call(cbind, sapply(ss, "[", 7)),do.call(cbind, sapply(ss, "[", 2)))
  id.size val.size       X1    X2
1       3        2 5.333333 11.50
2       4        4 2.250000 17.25
3       1        2 8.000000 13.50

[trinker]

Not sure if this is what you're after or if you're looking to automate this somehow:

df = data.frame(id=c(3,5,1,6,8,2,3,5), val=c(14,17,11,19,12,13,17,16)) 
ss=apply(df, 2, function(x) kmeans(x,3)) 
dat <- data.frame(do.call(cbind, sapply(ss, "[", 7)),do.call(cbind, sapply(ss, "[", 2))) 
names(dat)[3:4] <-c("new.name1", "new.name2")
dat

[masarimk]

Thank you Sir. I really appreciate that.