It is now supported by the MATLAB editor (as of R2022a).
The old ASCII notation is left in some files that we copy as-is from other
sources (e.g. in the contrib/ and m4/ subdirectories).
The particles submodule is not updated at this point, because it is in an
inconsistent state.
[skip ci]
The tests kalman_filter_smoother/fs2000_smoother_only.mod and
kalman_filter_smoother/fs2000_smoother_only_ns.mod would fail under MATLAB
R2014a.
They both run the estimation command without having declared an estim_params
block.
Their execution would fail in check_bounds_and_definiteness_estimation, at the
point where we check if initial values are below the lower bound.
The problem is that xparam1 would be of size 0×1, while bounds.lb would be
empty (0×0), and the comparison fails on older MATLABs.
The fix consists in ensuring that xparam1 remains of size 0×0 in that case.
Default value is 1 (initialization with the ergodic variance of the reduced
form solution of the model approximated at order one).
If the model has unit roots, the user must use `nonlinear_filter_initialization=3`,
which select an identity matrix for the initial covariance matrix of the state variables.
A side effect of this option is to temporarily change the value of options_.qz_criterium to
a value above one (ie 1+1e-6) so that the unit roots are not rejected. If the
model has unit roots and if the and if the option
nonlinear_filter_initialization has a value less than 3, the evaluation of the
likelihood will fail, because by default the unit root is counted as an
unstable root.
* Second and third positional arguments after the name of the
estimated parameter in the estimated_params block are only
considered in the optimization stage (not in the MCMC)
* Do not store bounds in bayestopt_, because bounds do not always
reflect restrictions implied by prior shapes.
* prior_bounds routine returns a structure (with fields lb and ub)
instead of a matrix.
(1) The non_linear_dsge_likelihood routine was not adapted to the last changes related to the way data are handled.
(2) The removal of the trend was missing.
Uses preprocessing capabilities introduced in 07137e804bFixes#392 and #494. Also fixes a bug in the checking for positive definiteness of covariance matrices in likelihood functions
Allows for calibrated covariances by reading them out and setting them after covariance matrix has been reconstructed from correlation and variances.
Adds unit test
Sébastien Villemot
Johannes Pfeifer
Stéphane Adjemian (Charybdis)
Houtan Bastani
Johannes Pfeifer
Michel Juillard