|
TROLL provides a variety of standard and experimental techniques for data analysis and parameter estimation, including the following:
- Single Equation Estimation
- Ordinary Least Squares
- Belsley-Kuh-Welsch Regression Diagnostics
- Cochrane-Orcutt Serial Correlation Correction
- Hildreth-Lu Serial Correlation Correction
- Polynomial Distributed Lags
- Non-Linear Least Squares
- Two-Stage Least Squares
- Equation Systems Estimation
- Three-Stage Least Squares
- Zellner's "Seemingly Unrelated Regressions"
- Full-Information Maximum Likelihood
- State Space Methods
- Kalman Filtering
- Optimal Smoothing
- Maximum Likelihood Hyperparameter Estimation
- Support for User-Defined Estimation Techniques
- Function Minimization
- Matrix Algebra
- Probability Functions
- Statitistical Functions
- Additional Data Analysis Support
- Seasonal Adjustment
- Standard Summary Statistics
- Correlations and Covariances
- Moving Averages
- Periodicity Conversion
- Spline-based Interpolation
\
| 
