Skip to content

Creating subsets of a matrix

4 messages · poolmunch, Joshua Wiley, jim holtman

poolmunch
# 
Hello, 

Say I have 2 columns, bmi and gender, the first being all the values and the
second being male or female. How would I subset this into males only and
females only? I have searched these fora and read endlessly about select[]
and split() functions but to no avail. Also the table is not ordered.

     bmi gender        ->      bmi gender         +    bmi  gender
1  24.78   male              1 24.78   male           3 23.18  female
2  26.42   male              2 26.42   male           4 22.36  female
3  23.18 female              ...                          ...
4  22.36 female
...

Thank you in advance
Joshua Wiley
# 
Hi,

Try looking at ?subset

I think something like this should work:

dat.male <- subset(dat, gender == "male")
dat.female <- subset(dat, gender == "female")

though it may require tweaking depending how your data is stored.  If
you want more specific help, suppose your matrix is called "dat" (but
substitute the actual name of it in), send us (or me) the output from:

dput(dat)
or if that is incredibly long, just
str(dat)

Here is a little example with a builtin dataset using subset():

print(mtcars) # show the data

## save and show all rows where column 'vs' is equal to 1
(dat.1 <- subset(mtcars, vs == 1))
## save and show all rows where column 'vs' is equal to 0
(dat.0 <- subset(mtcars, vs == 0))

Hope this helps,

Josh
On Sun, Jan 23, 2011 at 10:36 AM, poolmunch <poolmunch at gmail.com> wrote:

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      jim holtman
    

    

      
      #
    

  


  

      
Try 'split'

            

      
      
      
    

        
bmi gender
1 24.78   male
2 26.42   male
3 23.18 female
4 22.36 female

            

      
      
      
    

        
$female
    bmi gender
3 23.18 female
4 22.36 female

$male
    bmi gender
1 24.78   male
2 26.42   male

            

      
      
      
    

        

        
On Sun, Jan 23, 2011 at 1:36 PM, poolmunch <poolmunch at gmail.com> wrote:

            

      
      
      
    

        

      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      poolmunch
    

    

      
      #
    

  


  

      
# 4 . (a) Create subvectors, one for boys and one for girls.


bmi.male <- subset(bmi, gender == "male")	    # Creates a subset of bmi with
male only values n=30
bmi.female <- subset(bmi, gender == "female")	    # Creates a subset of bmi
with female only values n=30
male.bmi.only <-(bmi.male [, 1])			    # To help with the next questions I
remove the "male" and 
female.bmi.only <-(bmi.female [, 1])		    # "female" columns in favour of
naming the data sets.


I went with Josh's answer and I thank you for the help I received so
quickly.