Time Series Analysis and Examples

Overview

This chapter describes SAS/IML subroutines related to univariate, multivariate, and fractional time series analysis, and subroutines for Kalman filtering and smoothing. These subroutines can be used in analyzing economic and financial time series. You can develop a model of univariate time series and a model of the relationships between vector time series. The Kalman filter subroutines provide analysis of various time series and are presented as a tool for dealing with state space models.

The subroutines offer the following functions:

generating univariate, multivariate, and fractional time series
computing likelihood function of ARMA, VARMA, and ARFIMA models
computing an autocovariance function of ARMA, VARMA, and ARFIMA models
checking the stationarity of ARMA and VARMA models
filtering and smoothing of time series models by using Kalman method
fitting AR, periodic AR, time-varying coefficient AR, VAR, and ARFIMA models
handling Bayesian seasonal adjustment model

In addition, decomposition analysis, forecast of an ARMA model, and fractionally differencing of the series are provided.

This chapter consists of five sections. The first section includes the ARMACOV and ARMALIK subroutines and ARMASIM function. The second section includes the TSBAYSEA, TSDECOMP, TSMLOCAR, TSMLOMAR, TSMULMAR, TSPERARS, TSPRED, TSROOT, TSTVCAR, and TSUNIMAR subroutines. The third section includes the KALCVF, KALCVS, KALDFF, and KALDFS subroutines. The fourth section includes the VARMACOV, VARMALIK, VARMASIM, VNORMAL, and VTSROOT subroutines. The last section includes the FARMACOV, FARMAFIT, FARMALIK, FARMASIM, and FDIF subroutines.

Top of Page