2602 ✦ Chapter 37: The SASEHAVR Interface Engine Table 37.3 continued Database Name Offering Type Description REGIONW U.S. regional Selected regional indicators PIQR U.S. regional Personal income PIR U.S. regional Personal income PIRMSA U.S. regional Personal income PICOUNTY U.S. regional Personal income PIRC1 to 9 U.S. regional Personal income MBAMTG U.S. regional Mortgage delinquency rates from Mortgage Bankers Association DLINQR U.S. regional Consumer delinquency rates from American Bankers Association FALOAN U.S. regional Real estate and construction delinquency rates by Foresight Analytics BANKRUPT U.S. regional Bankruptcies by county and metropolitan statisti- cal area GSP U.S. regional Gross state product from BEA GDPMSA U.S. regional Gross domestic product by MSA USPOP U.S. regional Population by age and sex USPOPC U.S. regional Population by age and sex PORTS U.S. regional Trade by port EXPRQ1 to 9 U.S. regional Exports by industry and country from the World Institute for Strategic Economic Research and the U.S. Census Bureau EXPORTSR U.S. regional Exports by industry and country from the World Institute for Strategic Economic Research and the U.S. Census Bureau GOVFINR U.S. regional Government financial statistics from the U.S. Cen- sus Bureau and Rockefeller Institute of Govern- ment FDICR U.S. regional FDIC banking statistics References Haver Analytics (2009), DLX API Programmer’s Reference, New York [http://www.haver.com/] Haver Analytics (2009), DLX Database Profile, New York Haver Analytics (2009), Data Link Express, Time Series Data Base Management System, New York [http://www.haver.com/] Acknowledgments ✦ 2603 Acknowledgments Many people have been instrumental in the development of the ETS Interface engine. The individuals listed here have been especially helpful. Maurine Haver, Haver Analytics, New York, NY Lai Cheng, Haver Analytics, New York, NY Rick Langston, SAS Institute, Cary, NC The final responsibility for the SAS System lies with SAS alone. We hope that you will always let us know your opinions about the SAS System and its documentation. It is through your participation that SAS software is continuously improved. 2604 Part IV Time Series Forecasting System 2606 Chapter 38 Overview of the Time Series Forecasting System Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2607 Using the Time Series Forecasting System . . . . . . . . . . . . . . . . . . . . . . 2608 SAS Software Products Needed . . . . . . . . . . . . . . . . . . . . . . . . . . . 2609 Introduction The Time Series Forecasting system forecasts future values of time series variables by extrapolating trends and patterns in the past values of the series or by extrapolating the effect of other variables on the series. The system provides convenient point-and-click windows to control the time series analysis and forecasting tools of SAS/ETS software. You can use the system in a fully automatic mode, or you can use the system’s diagnostic features and time series modeling tools interactively to develop forecasting models customized to best predict your time series. The system provides both graphical and statistical features to help you choose the best forecasting method for each series. The following is a brief summary of the features of the Time Series Forecasting system. You can use the system in the following ways: use a wide variety of forecasting methods, including several kinds of exponential smoothing models, Winters method, and ARIMA (Box-Jenkins) models. You can also produce forecasts by combining the forecasts from several models. use predictor variables in forecasting models. Forecasting models can include time trend curves, regressors, intervention effects (dummy variables), adjustments you specify, and dynamic regression (transfer function) models. view plots of the data, predicted versus actual values, prediction errors, and forecasts with confidence limits, as well as autocorrelations and results of white noise and stationarity tests. Any of these plots can be zoomed and can represent raw or transformed series. use hold-out samples to select the best forecasting method 2608 ✦ Chapter 38: Overview of the Time Series Forecasting System compare goodness-of-fit measures for any two forecasting models side by side or list all models sorted by a particular fit statistic view the predictions and errors for each model in a spreadsheet or compare the fit of any two models in a spreadsheet examine the fitted parameters of each forecasting model and their statistical significance control the automatic model selection process: the set of forecasting models considered, the goodness-of-fit measure used to select the best model, and the time period used to fit and evaluate models customize the system by adding forecasting models for the automatic model selection process and for point-and-click manual selection save your work in a project catalog print an audit trail of the forecasting process show source statements for PROC ARIMA code save and print system output including spreadsheets and graphs Using the Time Series Forecasting System Chapters starting from Chapter 39, “Getting Started with Time Series Forecasting,” through Chap- ter 43, “Using Predictor Variables,” contain a series of example sessions that show the major features of the system. Chapters from Chapter 44, “Command Reference,” through Chapter 46, “Forecasting Process Details,” serve as reference and provide more details about how the system operates. The reference chapters contain a complete list of system features. To get started using the Time Series Forecasting system, it is a good idea to work through a few of the example sessions. Start with Chapter 39, “Getting Started with Time Series Forecasting,” and use the system to reproduce the steps shown in the examples. Continue with the other chapters when you feel comfortable using the system. The example sessions make use of time series data sets contained in the SASHELP li- brary: air, citimon, citiqtr, citiyr, citiwk, citiday, gnp, retail, usecon , and workers . You can use these data sets to work through the example sessions or to experiment further with the system. Once you are familiar with how the system operates, start working with your own data to build your own forecasting models. When you have questions, consult the reference chapters mentioned above for more information about particular features. The Time Series Forecasting system forecasts time series, that is, variables that consist of ordered observations taken at regular intervals over time. Since the Time Series Forecasting system is a part of the SAS software system, time series values must be stored as variables in a SAS data set or data SAS Software Products Needed ✦ 2609 view, with the observations representing the time periods. The data can also be stored in an external spreadsheet or data base if you license SAS/ACCESS software. The Time Series Forecasting System chapters refer to series and variables. Since time series are stored as variables in SAS data sets or data views, these terms are used interchangeably. However, the term series is preferred when attention is focused on the sequence of data values, and the term variable is preferred when attention is focused on the data set. SAS Software Products Needed The Time Series Forecasting system is part of SAS/ETS software. To use it, you must have a license for SAS/ETS. To use the graphical display features of the system, you must also license SAS/GRAPH software. 2610 Chapter 39 Getting Started with Time Series Forecasting Contents The Time Series Forecasting Window . . . . . . . . . . . . . . . . . . . . . . . . 2612 Outline of the Forecasting Process . . . . . . . . . . . . . . . . . . . . . . . . . . . 2617 Specify the Input Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2617 Provide a Valid Time ID Variable . . . . . . . . . . . . . . . . . . . . . . . . 2617 Select and Fit a Forecasting Model for Each Series . . . . . . . . . . . . . . 2618 Produce the Forecasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2618 Save Your Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2618 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2619 The Input Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2619 The Data Set Selection Window . . . . . . . . . . . . . . . . . . . . . . . . 2619 Time Series Data Sets, ID Variables, and Time Intervals . . . . . . . . . . . 2623 Automatic Model Fitting Window . . . . . . . . . . . . . . . . . . . . . . . . . . 2624 Produce Forecasts Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2632 The Forecast Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2634 Forecasting Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2638 Saving and Restoring Project Information . . . . . . . . . . . . . . . . . . . 2640 Sharing Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2644 Develop Models Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2645 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2645 Fitting Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2648 Model List and Statistics of Fit . . . . . . . . . . . . . . . . . . . . . . . . 2653 Model Viewer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2655 Prediction Error Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2657 Autocorrelation Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2658 White Noise and Stationarity Plots . . . . . . . . . . . . . . . . . . . . . . 2659 Parameter Estimates Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2661 Statistics of Fit Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2662 Changing to a Different Model . . . . . . . . . . . . . . . . . . . . . . . . 2663 Forecasts and Confidence Limits Plots . . . . . . . . . . . . . . . . . . . . 2664 Data Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2665 Closing the Model Viewer . . . . . . . . . . . . . . . . . . . . . . . . . . . 2666 . statistics References Haver Analytics (20 09) , DLX API Programmer’s Reference, New York [http://www.haver.com/] Haver Analytics (20 09) , DLX Database Profile, New York Haver Analytics (20 09) , Data Link Express, Time. part of SAS/ETS software. To use it, you must have a license for SAS/ETS. To use the graphical display features of the system, you must also license SAS/GRAPH software. 2610 Chapter 39 Getting. . . . 26 19 The Input Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 19 The Data Set Selection Window . . . . . . . . . . . . . . . . . . . . . . . . 26 19 Time Series