[backcolor=rgba(255, 255, 255, 0.917969)]Dynare 4.3.0 版发布了
[backcolor=rgba(255, 255, 255, 0.917969)]新功能如下：
[backcolor=rgba(255, 255, 255, 0.917969)]Here is the list of the main user-visible changes:
[backcolor=rgba(255, 255, 255, 0.917969)]* New major algorithms:
[backcolor=rgba(255, 255, 255, 0.917969)]- Nonlinear estimation with a particle filter based on a second order
[backcolor=rgba(255, 255, 255, 0.917969)]  approximation of the model, as in Fernández-Villaverde and Rubio-Ramírez
[backcolor=rgba(255, 255, 255, 0.917969)]  (2005); this is triggered by setting `order=2' in the `estimation' command
[backcolor=rgba(255, 255, 255, 0.917969)]- Extended path solution method as in Fair and Taylor (1983); see the
[backcolor=rgba(255, 255, 255, 0.917969)]  `extended_path' command
[backcolor=rgba(255, 255, 255, 0.917969)]- Support for Markov-Switching Structural Bayesian VARs (MS-SBVAR) along the
[backcolor=rgba(255, 255, 255, 0.917969)]  lines of Sims, Waggoner and Zha (2008) (see the dedicated section in the
[backcolor=rgba(255, 255, 255, 0.917969)]  reference manual)
[backcolor=rgba(255, 255, 255, 0.917969)]- Optimal policy under discretion along the lines of Dennis (2007); see the
[backcolor=rgba(255, 255, 255, 0.917969)]  `discretionary_policy' command
[backcolor=rgba(255, 255, 255, 0.917969)]- Identification analysis along the lines of Iskrev (2010); see the
[backcolor=rgba(255, 255, 255, 0.917969)]  `identification' command
[backcolor=rgba(255, 255, 255, 0.917969)]- The Global Sensitivity Analysis toolbox (Ratto, 2008) is now part of the
[backcolor=rgba(255, 255, 255, 0.917969)]  official Dynare distribution

[backcolor=rgba(255, 255, 255, 0.917969)]* Other algorithmic improvements:
[backcolor=rgba(255, 255, 255, 0.917969)]- Stochastic simulation and estimation can benefit from block decomposition
[backcolor=rgba(255, 255, 255, 0.917969)]  (with the `block' option of `model'; only at 1st order)
[backcolor=rgba(255, 255, 255, 0.917969)]- Possibility of running smoother and filter on a calibrated model; see the
[backcolor=rgba(255, 255, 255, 0.917969)]  `calib_smoother' command
[backcolor=rgba(255, 255, 255, 0.917969)]- Possibility of doing conditional forecast on a calibrated model; see the
[backcolor=rgba(255, 255, 255, 0.917969)]  `parameter_set=calibration' option of the `conditional_forecast' command
[backcolor=rgba(255, 255, 255, 0.917969)]- The default algorithm for deterministic simulations has changed and is now
[backcolor=rgba(255, 255, 255, 0.917969)]  based on sparse matrices; the historical algorithm (Laffargue, Boucekkine
[backcolor=rgba(255, 255, 255, 0.917969)]  and Juillard) is still available under the `stack_solve_algo=6'option of the
[backcolor=rgba(255, 255, 255, 0.917969)]  `simul' command
[backcolor=rgba(255, 255, 255, 0.917969)]- Possibility of using an analytic gradient for the estimation; see the
[backcolor=rgba(255, 255, 255, 0.917969)]  `analytic_derivation' option of the `estimation' command
[backcolor=rgba(255, 255, 255, 0.917969)]- Implementation of the Nelder-Mead simplex based optimization routine for
[backcolor=rgba(255, 255, 255, 0.917969)]  computing the posterior mode; available under the `mode_compute=8' option of
[backcolor=rgba(255, 255, 255, 0.917969)]  the `estimation' command
[backcolor=rgba(255, 255, 255, 0.917969)]- Implementation of the CMA Evolution Strategy algorithm for computing the
[backcolor=rgba(255, 255, 255, 0.917969)]  posterior mode; available under the `mode_compute=9' option of the
[backcolor=rgba(255, 255, 255, 0.917969)]  `estimation' command
[backcolor=rgba(255, 255, 255, 0.917969)]- New solvers for Lyapunov equations which can accelerate the estimation of
[backcolor=rgba(255, 255, 255, 0.917969)]  large models; see the `lyapunov' option of the `estimation' command
[backcolor=rgba(255, 255, 255, 0.917969)]- New solvers for Sylvester equations which can accelerate the resolution of
[backcolor=rgba(255, 255, 255, 0.917969)]  large models with block decomposition; see the `sylvester' option of the
[backcolor=rgba(255, 255, 255, 0.917969)]  `stoch_simul' and `estimation' commands
[backcolor=rgba(255, 255, 255, 0.917969)]- The `ramsey_policy' command now displays the planner objective value
[backcolor=rgba(255, 255, 255, 0.917969)]  function under Ramsey policy and stores it in `oo_.planner_objective_value'
[backcolor=rgba(255, 255, 255, 0.917969)]- Theoretical autocovariances are now computed when the `block' option is
[backcolor=rgba(255, 255, 255, 0.917969)]  present
[backcolor=rgba(255, 255, 255, 0.917969)]- The `linear' option is now compatible with the `block' and `bytecode'
[backcolor=rgba(255, 255, 255, 0.917969)]  options
[backcolor=rgba(255, 255, 255, 0.917969)]- The `loglinear' option now works with purely backward or forward models at
[backcolor=rgba(255, 255, 255, 0.917969)]  first order

[backcolor=rgba(255, 255, 255, 0.917969)]