Determine the middle of a distribution.

Note:   If you are running a release of SAS prior to SAS 9.0, you can determine the median value of a group by using the ORDINAL function to mimic the MEDIAN function.

The syntax of the ORDINAL function is :

    ORDINAL(count,argument,argument, . . .) ;

The ORDINAL function returns the count-th largest of the arguments, beginning with the lowest value. For instance, if you specify 4 as the value of the count parameter, ORDINAL will return the 4th largest value in the series of variables or values in the argument list.

With an odd number of variables, the median will be the single value that falls at the midpoint of the distribution. See Sample 1 on the Full Code tab.

With an even number of variables, there is no single middle value, therefore the median will be the average of the two middle values in the ordered distribution. See Sample 2 on the Full Code tab.

---

These sample files and code examples are provided by SAS Institute Inc. "as is" without warranty of any kind, either express or implied, including but not limited to the implied warranties of merchantability and fitness for a particular purpose. Recipients acknowledge and agree that SAS Institute shall not be liable for any damages whatsoever arising out of their use of this material. In addition, SAS Institute will provide no support for the materials contained herein.

---

/* Sample 1: Odd number of variables in the distribution */

data test;
   array x{9} x1-x9;

   /* Generate 9 variables, 5 observations */
   do i=1 to 5;
     do j=1 to 9;
       x{j}=ceil(ranuni(123)*10);
     end;
     output;
   end;
  drop i j;
run;


data test;
  set test; 
  median=ordinal(5,x1,x2,x3,x4,x5,x6,x7,x8,x9);
run;

proc print data=test; 
run;


/* Sample 2: Even number of variables in the distribution */

data test2;
  array x{10} x1-x10;

  /* Generate 10 variables, 5 observations */
  do i=1 to 5;
    do j=1 to 10;
      x{j}=ceil(ranuni(123)*10);
    end;
    output;
  end;
  drop i j;
run;


data test2;
  set test2;
  median=(ordinal(5,x1,x2,x3,x4,x5,x6,x7,x8,x9,x10)
         + ordinal(6,x1,x2,x3,x4,x5,x6,x7,x8,x9,x10))/2;
run;


proc print data=test2; 
run;

---

These sample files and code examples are provided by SAS Institute Inc. "as is" without warranty of any kind, either express or implied, including but not limited to the implied warranties of merchantability and fitness for a particular purpose. Recipients acknowledge and agree that SAS Institute shall not be liable for any damages whatsoever arising out of their use of this material. In addition, SAS Institute will provide no support for the materials contained herein.

Sample 1: Odd number of variables in the distribution 


Obs    x1    x2    x3    x4    x5    x6    x7    x8    x9    median

 1     8     4     2    10     4     3     8     4     2       4
 2     2     8     5    10     3     8     6     6     9       6
 3     2     9     7     8     8     4     6    10     1       7
 4     7     7     4     6     4     6     3     6     6       6
 5     1     2     6     7     5     2     8    10     8       6




Sample 2: Even number of variables in the distribution 


 Obs    x1    x2    x3    x4    x5    x6    x7    x8    x9    x10    median

  1     8     4     2    10     4     3     8     4     2     2       4.0
  2     8     5    10     3     8     6     6     9     2     9       7.0
  3     7     8     8     4     6    10     1     7     7     4       7.0
  4     6     4     6     3     6     6     1     2     6     7       6.0
  5     5     2     8    10     8    10     1     7     9     3       7.5

Determine the middle of a distribution.

Type:	Sample
Topic:	SAS Reference ==> DATA Step SAS Reference ==> Functions

Date Modified:	2005-12-23 03:02:57
Date Created:	2004-09-30 14:09:03

Operating System and Release Information

Product Family	Product	Host	SAS Release
			Starting	Ending
SAS System	Base SAS	All	n/a	n/a