The ARIMA Procedure
- Overview
-
Getting Started
The Three Stages of ARIMA ModelingIdentification StageEstimation and Diagnostic Checking StageForecasting StageUsing ARIMA Procedure StatementsGeneral Notation for ARIMA ModelsStationarityDifferencingSubset, Seasonal, and Factored ARMA ModelsInput Variables and Regression with ARMA ErrorsIntervention Models and Interrupted Time SeriesRational Transfer Functions and Distributed Lag ModelsForecasting with Input VariablesData Requirements -
Syntax
-
DetailsThe Inverse Autocorrelation FunctionThe Partial Autocorrelation FunctionThe Cross-Correlation FunctionThe ESACF MethodThe MINIC MethodThe SCAN MethodStationarity TestsPrewhiteningIdentifying Transfer Function ModelsMissing Values and AutocorrelationsEstimation DetailsSpecifying Inputs and Transfer FunctionsInitial ValuesStationarity and InvertibilityNaming of Model ParametersMissing Values and Estimation and ForecastingForecasting DetailsForecasting Log Transformed DataSpecifying Series PeriodicityDetecting OutliersOUT= Data SetOUTCOV= Data SetOUTEST= Data SetOUTMODEL= SAS Data SetOUTSTAT= Data SetPrinted OutputODS Table NamesStatistical Graphics
-
Examples
- References
The output data set produced by the OUT= option of the PROC ARIMA or FORECAST statements contains the following:
-
the BY variables
-
the ID variable
-
the variable specified by the VAR= option in the IDENTIFY statement, which contains the actual values of the response series
-
FORECAST, a numeric variable that contains the one-step-ahead predicted values and the multistep forecasts
-
STD, a numeric variable that contains the standard errors of the forecasts
-
a numeric variable that contains the lower confidence limits of the forecast. This variable is named L95 by default but has a different name if the ALPHA= option specifies a different size for the confidence limits.
-
RESIDUAL, a numeric variable that contains the differences between actual and forecast values
-
a numeric variable that contains the upper confidence limits of the forecast. This variable is named U95 by default but has a different name if the ALPHA= option specifies a different size for the confidence limits.
The ID variable, the BY variables, and the response variable are the only ones copied from the input to the output data set. In particular, the input variables are not copied to the OUT= data set.
Unless the NOOUTALL option is specified, the data set contains the whole time series. The FORECAST variable has the one-step forecasts (predicted values) for the input periods, followed by n forecast values, where n is the LEAD= value. The actual and RESIDUAL values are missing beyond the end of the series.
If you specify the same OUT= data set in different FORECAST statements, the latter FORECAST statements overwrite the output from the previous FORECAST statements. If you want to combine the forecasts from different FORECAST statements in the same output data set, specify the OUT= option once in the PROC ARIMA statement and omit the OUT= option in the FORECAST statements.
When a global output data set is created by the OUT= option in the PROC ARIMA statement, the variables in the OUT= data set are defined by the first FORECAST statement that is executed. The results of subsequent FORECAST statements are vertically concatenated onto the OUT= data set. Thus, if no ID variable is specified in the first FORECAST statement that is executed, no ID variable appears in the output data set, even if one is specified in a later FORECAST statement. If an ID variable is specified in the first FORECAST statement that is executed but not in a later FORECAST statement, the value of the ID variable is the same as the last value processed for the ID variable for all observations created by the later FORECAST statement. Furthermore, even if the response variable changes in subsequent FORECAST statements, the response variable name in the output data set is that of the first response variable analyzed.