Table of Contents
In order to calculate percentiles in SAS, one must first organize the data into sorted ascending order, assign a value to each data point, and then compute the percentile for each data point based on the percentage of the total number of data points it is greater than. Finally, you can use the SAS Percentile Procedure to generate the percentile values.
Here are the three most common ways to calculate the percentiles of a dataset in SAS:
Method 1: Calculate One Specific Percentile Value
/*calculate 70th percentile value for var1*/
proc univariate data=original_data;
var var1;
output out=percentile_data
pctlpts = 70
pctlpre = P_;
run;
Method 2: Calculate Multiple Specific Percentile Values
/*calculate 70th, 80th, and 90th percentile value for var1*/
proc univariate data=original_data;
var var1;
output out=percentile_data
pctlpts = 70 80 90
pctlpre = P_;
run;
Method 3: Calculate Percentiles by Group
/*sort original data by var2*/
proc sort data=original_data;
by var2;
run;
/*calculate percentiles for var1 grouped by var2*/
proc univariate data=original_data;
var var1;
by var2;
output out=percentile_data
pctlpts = 70, 80, 90
pctlpre = P_;
run;
Note: The pctlpts statement specifies which percentiles to calculate and the pctlpre statement specifies the prefix to use for the percentiles in the output.
The following examples show how to use each method with the following dataset in SAS:
/*create dataset*/
data original_data;
input team $ points;
datalines;
A 12
A 15
A 16
A 21
A 22
A 25
A 29
A 31
B 16
B 22
B 25
B 29
B 30
B 31
B 33
B 38
;
run;
/*view dataset*/
proc print data=original_data;
Example 1: Calculate One Specific Percentile Value
The following code shows how to calculate the 70th percentile for the points variable:
/*calculate 70th percentile value for points*/
proc univariate data=original_data;
var points;
output out=percentile_data
pctlpts = 70
pctlpre = P_;
run;
/*view results*/
proc print data=percentile_data;
The value at the 70th percentile turns out to be 30.
Example 2: Calculate Multiple Specific Percentile Values
/*calculate 70th, 80th, and 90th percentile value for points*/
proc univariate data=original_data;
var points;
output out=percentile_data
pctlpts = 70 80 90
pctlpre = P_;
run;
Here’s how to interpret the output:
- The value at the 70th percentile is 30.
- The value at the 80th percentile is 31.
- The value at the 90th percentile is 33.
Example 3: Calculate Percentiles by Group
The following code shows how to calculate the values at the 70th, 80th, 90th, and 95th percentile for the points variable, grouped by the team variable:
/*sort original data by team*/
proc sort data=original_data;
by team;
run;
/*calculate percentiles for points grouped by team*/
proc univariate data=original_data;
var points;
by team;
output out=percentile_data
pctlpts = 70, 80, 90 95
pctlpre = P_;
run;
The output table shows the values for the 70th, 80th, 90th, and 95th percentile for the points variable for both teams A and B.
The following tutorials explain how to perform other common tasks in SAS: