Output Data Sets
The OUT= data set saves the information displayed on a
Pareto chart.
If you specify CLASS= variables,
the OUT= data set contains one block of observations
for each combination of levels of the CLASS= variables,
and within each block there is an observation.
The observations are sorted in the order
in which the categories are displayed on the chart.
The following variables read from a DATA= data set are saved in
an OUT= data set:
- process variables
- CLASS= variables
- BY variables
- WEIGHT= variables
- the CTILES= variable
- the TILELEGEND= variable
- the NLEGEND= variable
- CBARS= variables
- PBARS= variables
- BARLEGEND= variables
In addition, the OUT= data set contains the following variables
that are created during the analysis:
- _COUNT_, which saves the frequency count
for each Pareto category
- _WCOUNT_, which saves the weighted
count for each category. This variable is created only
when you specify the WEIGHT= option.
- _PCT_, which saves the percent of the total
count for each category.
If you specify the WEIGHT= option, the variable _PCT_
saves the percent of the total weighted count.
- _CMPCT_, which saves the cumulative percent
for each category
See Output 29.8.2 for an example of an OUT= data set.
If you specify the MAXNCAT=, MAXCMPCT=, or MINPCT= option,
the OUT= data set saves only the
categories displayed on the chart. If
you create an OTHER= category
that merges the remaining categories, an additional
observation is saved with the new category.
Since the OTHER= value is defined as a formatted
value of the process variable, you should also
specify a corresponding internal value, as follows:
- If the process variable is a character variable,
specify the internal value with the OTHERCVAL= option.
If you do not specify this value, the OTHER=
value is saved as the internal value.
- If the process variable is a numeric variable,
specify the internal value with the OTHERNVAL= option.
If you do not specify this value, an internal
missing value is saved.
Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.