RATS (Regression Analysis of Time Series)

RATS for UNIX in QED is version 4.30 and for DOS in DUN 211 / DUN 350 is version 4.20. RATS is a time series software package for econometrics. The program can be used for any of the following tasks:

• Manage data sets and do a variety of data transformations.
• Applied Econometrics: Heteroscedasticity and Weighted Least Squares, Instrumental Variables Estimation, Two Stage Least Squares Regression, Nonlinear Least Squares Regression, Univariate or Multivariate Generalized Methods of Moments Estimation, Gerneralized Least Square, Seemingly Unrelated Regressions (SUR), Maximum-Likelihood (including ARCH, GARCH and related models).
• Hypothesis Testings: Granger and Sims Causality/Exogeneity Test, Hausman Test,etc.
• Nonstationarity Analysis: Unit Roots Tests (including Augmented Dickey-Fuller and Phillips-Perron tests), Cointegration Tests and Estimation (including Engle-Granger and Johansen).
• Vector Autoregression Analysis
• Forecasting, Simulations, Bootstrapping, and Monte Carlo Simulations
• Spectral Analysis (Frequency Domain)
• Cross Section and Panel Data
• Kalman Filtering
• Baisc Time Series Analysis including estimating ARIMA models, transfer function and intervention models.
• Logit and Probit Analysis
• Create, save, print, and export high-quality time series graphs and scatter plots.
• The most recent UNIX RATS (Version 4.3) that is only available at QED has additional new features: Linear and Quadratic Programming, Estimating a Neural Network and Generating Output from a Neural Net, and a new function to solve systems of linear equations.

TSP (Time Series Processor)

TSP for UNIX, LINUX in QED and for DOS in DUN 211 / DUN 350 are version 4.30A. TSP is a general purpose computer language for econometric and statistical data processing and estimation. The program can be used for any of the following tasks:

• Applied Econometrics
• Macroeconomic Research and Forecasting
• Monte Carlo Simulation
• Estimation and simulation of economic models
• All standard econometric techniques are available: ordinary least squares (OLS), two-stage least squares (instrumental variables, 2SLS), limited information maximum likelihood (LIML), polynomial and Shiller distributed lags, autoregressive correction, and weighted least squares.
• Advanced techniques are available: nonlinear least squares (NLS), estimation of GARCH models, Box-Jenkins estimation, multivariate regression, three stage least squares (3SLS), GMM, full information maximum likelihood (FIML), estimation with qualitative dependent variables (Probit/Logit), programmable maximum likelihood, and solution of linear and nonlinear system of equations.