See merge request !1869

See merge request !1868

In practice however, it seems that this bug never hits, because B and C always
have the same number of rows when the MEX is called.

This can make a difference when the return value of those function is directly
passed to a BLAS/LAPACK function.

On the other hand, if the return value is first stored in a pointer variable,
then it seems necessary to explicitly say that this pointer is also contiguous.

This avoids unnecessary array copies before calling BLAS/LAPACK functions.

This warns about hard-to-detect bugs where a function is called with an
implicit interface, thus possibly having mismatches in the calling sequence.

dynare_solve.m: have csolve honor selected options

See merge request !1867

Ref. !1865

Recommend that syntax in the manual, but keep the old one for backward
compatibility.

Ref. preprocessor#67

Ref. #569, !1863

evaluate_planner_objective.m: do not evaluate non-existing terms for linear models

Closes #1781

See merge request !1864

Closes Dynare/dynare#1781

Also add an integration test and fix homotopy error message.

Also removed commented codes.

Closes #1857

non_linear_dsge_likelihood.m: filter out explosive initial simulation

See merge request !1861

sim1_purely_backward.m: for purely static model, do not use past value for first iteration

See merge request !1862

Index 0 does not exist.

Implement new option smoother_redux, to allow fast smoother for very large...

See merge request !1859

prior_posterior_statistics_core.m: filter out cases where model does not solve

See merge request !1860

Marco Ratto authored 3 years ago and Johannes Pfeifer committed 3 years ago

Implement new option smoother_redux, to allow fast smoother for very large models. It runs smoother only for state variables and computes the remaining ones ex-post

Contains improvements, in order to recover as much as possible static unobserved (filtered, smoothed, updated, k-step ahead), Variance, State_uncertainty, k-step ahead variances trying to map lagged states onto current ones using pinv(T). This has exceptions (namely lagged shocks which are ONLY used to recover static NON observed variables). this exception is also trapped.
For such extensions we can only recover smoothed variables starting from d+1. Variances CANNOT be recovered for such variables (the smoother gives ZERO.)

Relax qz_criterium by factor of 10, when taking posterior subdraws. Also makes sure `prior` option works as intended.