The UNIVARIATE Procedure

Example 4.29 Adding a Distribution Reference Line

This example, which is a continuation of Example 4.28, illustrates how to add a reference line to a normal Q-Q plot, which represents the normal distribution with mean $\mu _0$ and standard deviation $\sigma _0$ . The following statements reproduce the Q-Q plot in Output 4.28.1 and add the reference line:

symbol v=plus;
title 'Normal Quantile-Quantile Plot for Hole Distance';
ods graphics off;
proc univariate data=Sheets noprint;
   qqplot Distance / normal(mu=est sigma=est color=red l=2)
                     square;
run;

The plot is displayed in Output 4.29.1.

Specifying MU=EST and SIGMA=EST with the NORMAL primary option requests the reference line for which $\mu _0$ and $\sigma _0$ are estimated by the sample mean and standard deviation. Alternatively, you can specify numeric values for $\mu _0$ and $\sigma _0$ with the MU= and SIGMA= secondary options. The COLOR= and L= options specify the color and type of the line, and the SQUARE option displays the plot in a square format. The NOPRINT options in the PROC UNIVARIATE statement and after the NORMAL option suppress all the tables of statistical output produced by default.

Output 4.29.1: Adding a Distribution Reference Line to a Q-Q Plot

The data clearly follow the line, which indicates that the distribution of the distances is normal.

A sample program for this example, uniex17.sas, is available in the SAS Sample Library for Base SAS software.