the error in reconstructing the k-step ahead forecasts with occbin smoother redux, should only occur in the first occbin smoother interation for non-stationary models (with diffuse filter), where tstart>1.

if k>tstart, the same simulation should have been done already in occbin.kalman_update, so it should never give an error

if k<=tstart, the simulation may crash, since we ignore OBC in the first (diffuse) steps and it may happen that, given the linear updated states, the occbin simulation does not converge.

Indeed, I never got a simulation failure in line 474, where this possibility is properly taken into account.

(cherry picked from commit abbd1118)

(cherry picked from commit 5ad067f1)

(cherry picked from commit 3e923ac9)

- trap the case when the occbin simulation does not converge (should only happen when first_period_occbin_update>1)

(cherry picked from commit 7041e107)

(cherry picked from commit 218a529c)

(cherry picked from commit 6ea07bf2)

Johannes Pfeifer authored 8 months ago and Stéphane Adjemian committed 8 months ago

(cherry picked from commit 895b5340)

Gregor Boehl authored 9 months ago and Sébastien Villemot committed 9 months ago

(cherry picked from commit dfa01edd)

[skip ci]

(cherry picked from commit 07191abf)

(cherry picked from commit d4211503)

(cherry picked from commit bf5ff6cf)

Normann Rion authored 11 months ago and Sébastien Villemot committed 10 months ago

(cherry picked from commit 7f62bf98)

[skip ci]

See the Debian changelog entry for dynare 6.1-2.

[skip ci]

(cherry picked from commit d1ffb60e)

See:
- https://github.com/Homebrew/homebrew-core/issues/165388
- https://trac.macports.org/ticket/69639

(manually cherry picked from commit 54956c22)

(manually cherry picked from commit 673ceb6d)

[skip ci]

(cherry picked from commit 0b8b05bc)

Johannes Pfeifer authored 11 months ago and Sébastien Villemot committed 10 months ago

(cherry picked from commit e22c3405)

Fixes https://forum.dynare.org/t/invalid-mex-file-whe-running-sensitivity-analysis-m1-mac-6-1/25829

It seems that meson does not link libquadmath (which is needed by libgfortran).
When doing dynamic linking, this is not an issue since the dynamic linker loads
libquadmath recursively after loading libgfortran. But when doing static linking,
libquadmath needs to be explicitly linked.

Not yet clear at this point: why isn’t this needed under Linux and Windows?

(cherry picked from commit f1e10960)

This avoids interfering with the build-matlab directory that may already be
used by the developper locally. Also, this makes it symmetric with the Windows
build script.

(cherry picked from commit 1e66daf7)

(cherry picked from commit a3f44cc6)

(manually cherry picked from commit 3256cf50)

These should be strict inequalities.

(cherry picked from commit 3cd6bd45)

Johannes Pfeifer authored 11 months ago and Sébastien Villemot committed 11 months ago

Crashed if mod()==0 as last value was not accounted for

(cherry picked from commit 70ebf0ab)

The solver relies on a reordering of equations (eq_index) so that reordered
equations appear at the index of the endogenous to which they are associated
through the complementarity condition.

However, this reordering could be computed incorrectly in the presence of
multiple complementarity conditions (technically, eq_index was not necessarily
a permutation).

(cherry picked from commit 1fea7f3e)

Johannes Pfeifer authored 11 months ago and Sébastien Villemot committed 11 months ago

homotopy expects a matrix conformable with the initval_series

(cherry picked from commit aeb1b92e)

Willi Mutschler authored 1 year ago and Sébastien Villemot committed 1 year ago

[skip ci]

(cherry picked from commit 8b7f2509)

[skip ci]

[skip ci]

(cherry picked from commit a9687a2d)

[skip ci]

(cherry picked from commit 35a8d230)

Willi Mutschler authored 1 year ago and Sébastien Villemot committed 1 year ago

(cherry picked from commit 1d86e8d2)

Johannes Pfeifer authored 1 year ago and Sébastien Villemot committed 1 year ago

Prevents crash

(cherry picked from commit b9d069da)

Johannes Pfeifer authored 1 year ago and Sébastien Villemot committed 1 year ago

Occurs with e.g. one shock only for heteroskedastic_filter

(cherry picked from commit 8b7420c7)

(manually cherry picked from commit 840d938f)

[skip ci]

(cherry picked from commit 240c0eb0)

Johannes Pfeifer authored 1 year ago and Sébastien Villemot committed 1 year ago

 disp_moments.m: fix crash in variance decomposition with measurement errors when zero variance shocks are present

(cherry picked from commit 8b703bb4)

Johannes Pfeifer authored 1 year ago and Sébastien Villemot committed 1 year ago

(cherry picked from commit 407def27)

(cherry picked from commit 25ffd847)

Normann Rion authored 1 year ago and Sébastien Villemot committed 1 year ago

Fix `simult_.m` to call specialized simulation codes for order <= 3 instead of `k_order_simul` whenever possible

(manually cherry picked from commit 03064501)