Navigation:  Release Notes >

 

YADA Version 3.10

 

Print this Topic Previous pageReturn to chapter overviewNext page

Released on 19/03/2012:

YADA now supports the univariate Kalman filter and smoother. The basic algorithms are described by Koopman and Durbin (2000).
Diffuse initialization of the Kalman filter has been added.
Monte Carlo filtering for investigating the sensitivity of the DSGE model solution to the values of different parameters. The approach is discussed in some detail by Ratto (2008).
YADA now calculates skewness and kurtosis from the DSGE and DSGE-VAR models when simulating the observed variables. For details, see observed variable correlations for the DSGE models.
Conditional forecasting now supports time-varying number of conditioning assumptions and lag-varying K2 matrices. The former are imposed by checking the conditioning assumptions for NaN values, and the latter through K2 being 3-dimensional. For details, see data construction file.
Sequential marginal likelihood estimation and eigenvalue plots for DSGE-VARs.
The Laplace approximation of the marginal likelihood of the DSGE model can now be recomputed for the full sample from the Sequential marginal likelihood estimation menu-item on the View menu.
It is now possible to restrict the posterior draws of the VAR parameters in DSGE-VARs and BVARs to imply a stationary VAR. This option is available on the lag order selection frame on the Bayesian VAR tab.

 

 


Page url: http://www.texlips.net/yada/help/index.html?3_10.htm