PROC CAPABILITY and General Statements
CLASS variable-1 <(v-options)> <variable-2 <(v-options)>></ KEYLEVEL= value1 | (value1 value2)> ;
The CLASS statement specifies one or two variables used to group the data into classification levels. Variables in a CLASS
statement are referred to as CLASS variables. CLASS variables can be numeric or character. Class variables can have floating point values, but they typically have a few
discrete values that define levels of the variable. You do not have to sort the data by CLASS variables. PROC CAPABILITY uses
the formatted values of the CLASS variables to determine the classification levels.
Note: You cannot specify a COMPHISTOGRAM statement together with a CLASS statement.
You can specify the following v-options enclosed in parentheses after a CLASS variable:
-
MISSING
-
specifies that missing values for the CLASS variable are to be treated as valid classification levels. Special missing values
that represent numeric values ('.A' through '.Z' and '._') are each considered as a separate value. If you omit MISSING, PROC
CAPABILITY excludes the observations with a missing CLASS variable value from the analysis. Enclose this option in parentheses
after the CLASS variable.
-
ORDER=DATA | FORMATTED | FREQ | INTERNAL
-
specifies the display order for the CLASS variable values. The default value is INTERNAL. You can specify the following values
with the ORDER= option:
- DATA
-
orders values according to their order in the input data set. When you use a plot statement, PROC CAPABILITY displays the
rows (columns) of the comparative plot from top to bottom (left to right) in the order that the CLASS variable values first
appear in the input data set.
- FORMATTED
-
orders values by their ascending formatted values. This order might depend on your operating environment. When you use a plot
statement, PROC CAPABILITY displays the rows (columns) of the comparative plot from top to bottom (left to right) in increasing
order of the formatted CLASS variable values. For example, suppose a numeric CLASS variable DAY (with values 1, 2, and 3)
has a user-defined format that assigns Wednesday to the value 1, Thursday to the value 2, and Friday to the value 3. The rows
of the comparative plot will appear in alphabetical order (Friday, Thursday, Wednesday) from top to bottom.
If there are two or more distinct internal values with the same formatted value, then PROC CAPABILITY determines the order
by the internal value that occurs first in the input data set. For numeric variables without an explicit format, the levels
are ordered by their internal values.
- FREQ
-
orders values by descending frequency count so that levels with the most observations are listed first. If two or more values
have the same frequency count, PROC CAPABILITY uses the formatted values to determine the order.
When you use a plot statement, PROC CAPABILITY displays the rows (columns) of the comparative plot from top to bottom (left
to right) in order of decreasing frequency count for the CLASS variable values.
- INTERNAL
-
orders values by their unformatted values, which yields the same order as PROC SORT. This order may depend on your operating
environment.
When you use a plot statement, PROC CAPABILITY displays the rows (columns) of the comparative plot from top to bottom (left
to right) in increasing order of the internal (unformatted) values of the CLASS variable. The first CLASS variable is used
to label the rows of the comparative plots (top to bottom). The second CLASS variable is used to label the columns of the
comparative plots (left to right). For example, suppose a numeric CLASS variable DAY (with values 1, 2, and 3) has a user-defined
format that assigns Wednesday to the value 1, Thursday to the value 2, and Friday to the value 3. The rows of the comparative
plot will appear in day-of-the-week order (Wednesday, Thursday, Friday) from top to bottom.
You can specify the following options after the slash (/) in the CLASS statement.
-
KEYLEVEL=value | ( value1 value2 )
-
specifies the key cells in comparative plots. For each plot, PROC CAPABILITY first determines the horizontal axis scaling for the key cell, and then
extends the axis using the established tick interval to accommodate the data ranges for the remaining cells, if necessary.
Thus, the choice of the key cell determines the uniform horizontal axis that PROC CAPABILITY uses for all cells.
If you specify only one CLASS variable and use a plot statement, KEYLEVEL=value identifies the key cell as the level for which the CLASS variable is equal to value. By default, PROC CAPABILITY sorts the levels in the order determined by the ORDER= option, and the key cell is the first
occurrence of a level in this order. The cells display in order from top to bottom or left to right. Consequently, the key
cell appears at the top (or left). When you specify a different key cell with the KEYLEVEL= option, this cell appears at the
top (or left).
If you specify two CLASS variables, use KEYLEVEL= ( value1 value2) to identify the key cell as the level for which CLASS variable n is equal to valuen. By default, PROC CAPABILITY sorts the levels of the first CLASS variable in the order that is determined by its ORDER= option.
Then, within each of these levels, it sorts the levels of the second CLASS variable in the order that is determined by its
ORDER= option. The default key cell is the first occurrence of a combination of levels for the two variables in this order.
The cells display in the order of the first CLASS variable from top to bottom and in the order of the second CLASS variable
from left to right. Consequently, the default key cell appears at the upper left corner. When you specify a different key
cell with the KEYLEVEL= option, this cell appears at the upper left corner.
The length of the KEYLEVEL= value cannot exceed 16 characters and you must specify a formatted value.
The KEYLEVEL= option has no effect unless you specify a plot statement.
-
NOKEYMOVE
-
specifies that the location of the key cell in a comparative plot be unchanged by the CLASS statement KEYLEVEL= option. By
default, the key cell is positioned as the first cell in a comparative plot.
The NOKEYMOVE option has no effect unless you specify a plot statement.
Copyright © SAS Institute Inc. All Rights Reserved.
