• support.sas.com

• Knowledge Base

• Support

• Training & Bookstore

• Community

Expressions

Expressions can be used for conditional processing and on plot statements to calculate values that are not in the data. For example, an expression could be used on a LINEPARM statement to calculate the slope of the line that the statement produces.

For conditional processing, an expression is specified in parentheses on an IF statement. The following example illustrates where to specify an expression in an IF statement:

IF( expression )
       ODS-Graphs-statement(s);
ENDIF;

For calculating values that are needed on a plot statement, an expression is specified as the argument of an EVAL() construct on an appropriate plot argument or option. The following example illustrates where to specify an expression in a plot statement:

SLOPE = EVAL( expression )

An expression can be formed with constants, data columns, dynamic variables, functions, or other expressions. The following example illustrates an expression in which the dynamic variables Age, Height, and Weight are used:

(...)
dynamic Age Height Weight;
       (...)
EVAL( Age * Height + Weight ** 2 )

The following example illustrates an expression in which the column variable Time and the SGE functions MEAN and ACF are used:

EVAL( MEAN( Time ) + ACF( Time, NLags=10 ))

The following example shows the general form of an expression:

operand < operator-1 operand-1 ... operator-n operand-n >

An operand can be any of the following values:

constant

specifies a number or a string value. String values must be specified in quotes.

data column

specifies a data column.

dynamic variable

specifies a symbol that references a value or a column that is provided by the procedure at the time the graph is being created.

expression

specifies a nested expression. To control processing, the nested expression can use parentheses to determine the order of operations.

functions

specifies a function in the DATA step; SGE function, computation supported by the Statistical Graphics Engine; or ODS graphics function. When names collide, the number of function arguments is used to clarify the context. If there is more than one argument, it is assumed to be a SAS DATA step function. The following ODS Graphics functions are available:

Autocorrelation Functions for Time Series Analysis

numeric-column= function-name (numeric-column, NLags= integer)
Function Name	Description
ACF	Autocorrelation Function
ACFSTD	Autocorrelation Function Standard Errors
ACFNORM	Autocorrelation Function Normalized Statistics
ACFPROB	Autocorrelation Function Significance Probabilities
ACFLPROB	Autocorrelation Function Log Significance Probabilities
IACF	Inverse Autocorrelation
IACFSTD	Inverse Autocorrelation Standard Errors
IACFNORM	Inverse Autocorrelation Normalized Statistics
IACFPROB	Inverse Autocorrelation Significance Probabilities
IACFLPROB	Inverse Autocorrelation Log Significance Probabilities
PACF	Partial Autocorrelation
PACFSTD	Partial Autocorrelation Standard Errors
PACFNORM	Partial Autocorrelation Normalized Statistics
PACFPROB	Partial Autocorrelation Significance Probabilities
PACFLPROB	Partial Autocorrelation Log Significance Probabilities
WN	White Noise Test
WNPROB	White Noise Significance Probabilities
WNLPROB	White Noise Log Significance Probabilities

Decomposition Functions for Forecasting and Time Series Analysis

numeric-column= function-name (numeric-column, Seasonality= integer)
Function Name	Description
CC	Cycle-Component
DCOMP	Seasonal Decomposition
IC	Irregular Component
SA	Seasonal Adjusted
SC	Seasonal Component
SIC	Seasonal-Irregular Component
TC	Trend Component
TCC	Trend-Cycle Component
TCS	Trend-Cycle-Seasonal Component

Miscellaneous Functions

integer= function-name( column )
Function Name	Description
EXISTS	1 if specified column exists, 0 otherwise

Sorting Functions

column= function-name( column )
Function Name	Description
ARANK	Ascending Rank
ASORT	Ascending Sort
DRANK	Descending Rank
DSORT	Descending Sort
INDEX	Ascending Indices
RANK	Ascending Rank
SORT	Ascending Sort

Statistical Functions

Statistical functions operate as the DATA step functions of the same name.

number= function-name( numeric-column )
Function Name	Description
CSS	Corrected Sum of Squares
CV	Coefficient of Variation
KURTOSIS	Kurtosis
LCLM	Confidence Limit Below the Mean
MAX	Largest (Maximum) Value
MEAN	Mean
MEDIAN	Median (50th Percentile)
MIN	Smallest (Minimum) Value
N	Number of Non-Missing Values
NMISS	Number of Missing Values
P1	First Percentile
P5	Fifth Percentile
P25	Twenty-fifth Percentile
P50	Fiftieth Percentile
P75	Seventy-fifth Percentile
P90	Ninetieth Percentile
P95	Ninety-fifth Percentile
P99	Ninety-ninth Percentile
PROBT	p-value for T Statistic of a Student
Q1	First Quartile
Q3	Third Quartile
QRANGE	Interquartile Range
RANGE	Range
SKEWNESS	Skewness
STDDEV	Standard Deviation
STDERR	Standard Error of the Mean
SUM	Sum
SUMWGT	Sum of Weights
T	T Statistic of a Student
UCLM	Confidence Limit Above the Mean
USS	Uncorrected Sum of Squares
VAR	Variance

operators

specifies a standard SAS WHERE clause operator, or operators.