qz_criterium = 1+1e-6
- stoch_simul
- osr
- check
- dynare_estimation_1 with lik_init =2 or 3
qz_criterium = 1+1e-6 for dynare_estimation_1 with lik_init = 1
This correct a bug in estimation and computation of the posterior distribution of moments of endogenous variables when the priors permit roots too close to 1.
(--with-openmp) is not passed by an environment variable anymore. The function set_dynare_threads changes the default value of the
number of threads (default is 1) in the options_.threads structure. Changed calls to sparse_hessian_times_B_kronecker_C and
A_times_B_kronecker_C dlls accordingly.
1) safe management of remote folder;
2) shh scp (and windows equivalent) commands are concentrated in the master;
3) management of remote crashes;
4) begin building option to interrupt remote processes;
5) better parallel waitbar: pops up only at STARTUP (can be minimized subsequently); adaptive dimension;
(2) Speed improvement in second order simulations (replaced call to matlab's kron function by call to
A_times_B_kronecker_C).
(3) Removed useless globals.
(4) Cosmetic changes and corrections on headers.
* give the possibility of calibrating measurement errors in the "shocks" blocks (only for observed endogenous variables)
* M_.H is now initialized in the preprocessor
* only one "varobs" statement is now accepted in a MOD file
- reverting part of last change to handling of Qi and Ri in ComputingTasks.cc: r is the number of lags present in the restrictions
- fixing handling of Dirichlet option in mhm_input.dat. This fixes the strange Signal messages when running C executables.
- fix path in test_upper_cholesky.mod
by numgrad routines to compute the gradient. The default value (defined in global_initialization.m) for this parameter is 1e-6. The
value of epsilon is stored in options_.gradient_epsilon.
git-svn-id: https://www.dynare.org/svn/dynare/trunk@3335 ac1d8469-bf42-47a9-8791-bf33cf982152
* removed "options_.simul", and instead test if "options_.periods" is non-zero
* test for the incompatibility of options "periods" and "hp_filter" in "stoch_simul.m", instead of in the preprocessor
Reference manual:
* removed "simul" options, updated "periods"
* updated "order" option (for 3rd order)
* added "k_order_solver" option
* give some hints for installing a compiler for users of MATLAB for Windows
git-svn-id: https://www.dynare.org/svn/dynare/trunk@3188 ac1d8469-bf42-47a9-8791-bf33cf982152
(the terminal condition is y_{T} = y^{\star}), other possible values
are 1 (terminal condition is y_{T+1}=y_{T}) and 2 (terminal
condition is y_{T+1}=TransitionMatrix*y_{T}, where TransitionMatrix
is given by the first order approximation of the reduced form model).
* Added mode options_.terminal_condition=2 in perfect_foresight_simulation.m.
git-svn-id: https://www.dynare.org/svn/dynare/trunk@3176 ac1d8469-bf42-47a9-8791-bf33cf982152