R help - Aug 2011 - How to order based on the second two columns?

Hello All, 

I have a data frame consisting of 4 columns (id1, id2, y, pred)
where pred is the predicted value based on the glm function and my data frame is
called "all". "data" is another data frame that has all data
but I want to put together some important columns from my original data frame
(data) into another data frame (all) as follows and I would like them to be
sorted based on the id1 and id2. Here is what I do:

all_data = cbind(oder(data[,2]), order(data[,3]), data[,1], pred)
all = as.data.frame(all_data)
colnames(all) = c("id1", "id2", "y" ,
"pred")

when I do the ordering thing, I am not sure why I do not get the corresponding
"y" and "pred" values for that specific row after ordering.
Am I doing something wrong in here?

Thanks a lot,
ANDRA

On Aug 29, 2011, at 2:40 PM, Andra Isan wrote:
> Hello All,
>
> I have a data frame consisting of 4 columns (id1, id2, y, pred)
> where pred is the predicted value based on the glm function and my  
> data frame is called "all". "data" is another data
frame that has
> all data but I want to put together some important columns from my  
> original data frame (data) into another data frame (all) as follows  
> and I would like them to be sorted based on the id1 and id2. Here is  
> what I do:
>
> all_data = cbind(oder(data[,2]), order(data[,3]), data[,1], pred)
> all = as.data.frame(all_data)
> colnames(all) = c("id1", "id2", "y" ,
"pred")
>
> when I do the ordering thing, I am not sure why I do not get the  
> corresponding "y" and "pred" values for that specific
row after
> ordering. Am I doing something wrong in here?
Your error is in not using 'merge' instead of 'cbind'.

-- 
David Winsemius, MD
West Hartford, CT