The UNIVARIATE Procedure

Example 4.7 Saving Summary Statistics in an OUT= Output Data Set

This example illustrates how to save summary statistics in an output data set. The following statements create a data set named Belts, which contains the breaking strengths (Strength) and widths (Width) of a sample of 50 automotive seat belts:

data Belts;
   label Strength = 'Breaking Strength (lb/in)'
         Width    = 'Width in Inches';
   input Strength Width @@;
   datalines;
1243.51  3.036  1221.95  2.995  1131.67  2.983  1129.70  3.019
1198.08  3.106  1273.31  2.947  1250.24  3.018  1225.47  2.980
1126.78  2.965  1174.62  3.033  1250.79  2.941  1216.75  3.037
1285.30  2.893  1214.14  3.035  1270.24  2.957  1249.55  2.958
1166.02  3.067  1278.85  3.037  1280.74  2.984  1201.96  3.002
1101.73  2.961  1165.79  3.075  1186.19  3.058  1124.46  2.929
1213.62  2.984  1213.93  3.029  1289.59  2.956  1208.27  3.029
1247.48  3.027  1284.34  3.073  1209.09  3.004  1146.78  3.061
1224.03  2.915  1200.43  2.974  1183.42  3.033  1195.66  2.995
1258.31  2.958  1136.05  3.022  1177.44  3.090  1246.13  3.022
1183.67  3.045  1206.50  3.024  1195.69  3.005  1223.49  2.971
1147.47  2.944  1171.76  3.005  1207.28  3.065  1131.33  2.984
1215.92  3.003  1202.17  3.058
;

The following statements produce two output data sets containing summary statistics:

proc univariate data=Belts noprint;
   var Strength Width;
   output out=Means         mean=StrengthMean WidthMean;
   output out=StrengthStats mean=StrengthMean std=StrengthSD 
                            min=StrengthMin   max=StrengthMax;
run;

When you specify an OUTPUT statement, you must also specify a VAR statement. You can use multiple OUTPUT statements with a single procedure statement. Each OUTPUT statement creates a new data set with the name specified by the OUT= option. In this example, two data sets, Means and StrengthStats, are created. See Output 4.7.1 for a listing of Means and Output 4.7.2 for a listing of StrengthStats.

Output 4.7.1: Listing of Output Data Set Means

Analysis of Speeding Data

Obs StrengthMean WidthMean
1 1205.75 3.00584



Output 4.7.2: Listing of Output Data Set StrengthStats

Analysis of Speeding Data

Obs StrengthMean StrengthSD StrengthMax StrengthMin
1 1205.75 48.3290 1289.59 1101.73



Summary statistics are saved in an output data set by specifying keyword=names after the OUT= option. In the preceding statements, the first OUTPUT statement specifies the keyword MEAN followed by the names StrengthMean and WidthMean. The second OUTPUT statement specifies the keywords MEAN, STD, MAX, and MIN, for which the names StrengthMean, StrengthSD, StrengthMax, and StrengthMin are given.

The keyword specifies the statistic to be saved in the output data set, and the names determine the names for the new variables. The first name listed after a keyword contains that statistic for the first variable listed in the VAR statement; the second name contains that statistic for the second variable in the VAR statement, and so on.

The data set Means contains the mean of Strength in a variable named StrengthMean and the mean of Width in a variable named WidthMean. The data set StrengthStats contains the mean, standard deviation, maximum value, and minimum value of Strength in the variables StrengthMean, StrengthSD, StrengthMax, and StrengthMin, respectively.

See the section OUT= Output Data Set in the OUTPUT Statement for more information about OUT= output data sets.

A sample program for this example, uniex06.sas, is available in the SAS Sample Library for Base SAS software.