Saving Summary Statistics

[See SHWMCHR in the SAS/QC Sample Library]In this example, the MCHART statement is used to create a summary data set that can be read later by the SHEWHART procedure (as in the preceding example). The following statements read measurements from the data set Detergent and create a summary data set named Dethist:

proc shewhart data=Detergent;
   mchart Weight*Lot / outhistory = Dethist
                       nochart;
run;

The OUTHISTORY= option names the output data set, and the NOCHART option suppresses the display of the chart, which would be identical to the chart in Figure 13.13.4. Figure 13.13.7 contains a partial listing of Dethist.

There are four variables in the data set Dethist.

• Lot contains the subgroup index.

• WeightM contains the subgroup medians.

• WeightR contains the subgroup ranges.

• WeightN contains the subgroup sample sizes.

Note that the summary statistic variables are named by adding the suffix characters M, R, and N to the process Weight specified in the MCHART statement. In other words, the variable naming convention for OUTHISTORY= data sets is the same as that for HISTORY= data sets.

If you specify the STDDEVIATIONS option, the OUTHISTORY= data set includes a subgroup standard deviation variable instead of a subgroup range variable, as demonstrated by the following statements:

proc shewhart data=Detergent;
   mchart Weight*Lot / outhistory = Dethist2
                       stddeviations
                       nochart;
run;

Figure 13.13.8 contains a partial listing of Dethist2.

The variable WeightS, which contains the subgroup standard deviations, is named by adding the suffix character S to the process Weight.

For more information, see OUTHISTORY= Data Set.