Skip to content

complex splits

6 messages · Tim Wilson, Pierre Kleiber, Thomas Lumley

Original
Tim Wilson
# 
Hi everyone,

I'm having trouble figuring out how to split a dataframe more than once.

Let's say I have a dataframe d with a certain column called splitcol
composed of four possible ordinal values.  The same dataframe has
two other columns, col1 and col2, that have one of two possible values
each. I'd like to split d$splitcol based on col1 and col2 so I can report
frequencies of the subgroups.

So if col1 and col2 can each have two different values, after splitting
I should have four vectors based on d$splitcol.

Can I do this in one step, or do I need to do some intermediate
splitting?

-Tim
Pierre Kleiber
# 
split(d$splitcol,paste(d$col1,d$col2))

will give you vectors as elements of a list based on possible
combinations of col1 and col2.  Is that what you're after?
   Cheers, Pierre
Tim Wilson wrote:

            

      
      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Tim Wilson
    

    

      
      #
    

  


  

      
On Thu, Aug 15, 2002 at 02:30:08PM -1000, Pierre Kleiber wrote:

            

      
      
      
    

        
No, when I try this I only get two groups based on d$col1. What I'm
looking for is a split on d$col1 (producing two groups) and a subsequent
split on d$col2 for a total of four groups. Then I'll do frequencies of
those four groups.

-Tim

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Pierre Kleiber
    

    

      
      #
    

  


  

      
Strange that you only get two groups.  Note that col2 is
pasted to col1.   Here's what I get:


 > d <- data.frame(col1=sample(c("a","b"),20,
         rep=T),col2=sample(c("c","d"),20,rep=T),
         splitcol=sample(LETTERS[1:4],20,rep=T))
 > d
    col1 col2 splitcol
1     a    c        A
2     b    d        B
3     a    c        B
4     b    c        C
5     a    d        A
6     a    d        C
7     b    c        D
8     a    d        D
9     b    c        B
10    a    c        D
11    b    c        C
12    b    d        C
13    b    d        B
14    a    d        C
15    a    c        B
16    a    d        A
17    b    d        C
18    b    d        A
19    b    c        A
20    b    d        C
 > split(d$splitcol,paste(d$col1,d$col2))
$"a c"
[1] A B D B
Levels:  A B C D

$"a d"
[1] A C D C A
Levels:  A B C D

$"b c"
[1] C D B C A
Levels:  A B C D

$"b d"
[1] B C B C A C
Levels:  A B C D

        
Tim Wilson wrote:

        

            

      
      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Tim Wilson
    

    

      
      #
    

  


  

      
On Thu, Aug 15, 2002 at 02:55:36PM -1000, Pierre Kleiber wrote:

            

      
      
      
    

        
Thanks Pierre! It works now. I must have typed something wrong before.

OK, everyone move along...nothing more to see here. :-)

-Tim

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Thomas Lumley
    

    

      
      #
    

  


  

      
On Thu, 15 Aug 2002, Pierre Kleiber wrote:

        

            

      
      
      
    

        
or more simply

	split(d$splitcol,list(d$col1,d$col2))

	-thomas


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._