What is TSP?
TSPTM is a complete language for the estimation and simulation of econometric models. It is a world-wide standard for econometric estimation, with over 2000 installations. What follows is a general overview; detailed lists of features and commands can be found elsewhere. TSP stands for "Time Series Processor", although it is also commonly used with cross section and panel data. TSP happens to be a common acronym with alternative definitions.
FEATURES
- Easy-to-use free format command and data input
- All the standard econometric estimation methods, such as OLS, instrumental variables, LIML, nonlinear systems, generalized methods of moments, FIML, maximum likelihood for qualitative dependent variable models, ARIMA, Kalman filter, ARCH, and other time series techniques (complete feature list).
- Extensive diagnostics and testing facilities.
- Flexible data transformation with many built-in functions and matrix algebra.
- Offers a choice between interactive use or full programming language for econometric methods development.
COMMON APPLICATIONS
- Applied econometrics
- Macroeconomic research and forecasting
- Sales forecasting
- Financial analysis
- Cost analysis and forecasting
- Monte Carlo simulation
- Estimation and simulation of economic models
Although TSP was originally and continues to be developed primarily by economists, there is nothing in its design limiting it to economic times series. Any data consisting of repeated observations of the same variable for different units may be analyzed with TSP.
Official website: http://www.tspintl.com/
[此贴子已经被作者于2006-7-25 3:48:16编辑过]