TODO:
– investigate and fix the multi-threading performance issues
– fix compilation failure on macOS
– add support for the Microsoft x64 ABI (for Windows)
— create a separate unit test (instead of hijacking an existing test)
– add a version with pruning?

[skip CI]

– Sparse representation: fix bug in output of indices in driver and JSON files
– Minor simplification in static block decomposition driver output

dynare_estimation_init.m: set observable mapping for inversion filter if only smoother is used

See merge request !2116

– Sparse representation: compatibility fix with MATLAB < R2020a (Closes: #1875)
– Sparse representation: fix bug in output of indices in driver and JSON files

Manual: document passing of strings to macro processor

Closes #1877

See merge request !2115

Closes #1877

Intigriti reference: DYNARE-7TULYVR8

prior_posterior_statistics.m: fix logical condition

See merge request !2114

Make default jscale depend on number of estimated parameters

See merge request Dynare/dynare!2041

Related to #1849

[skip ci]

Use the new time-recursive block decomposition computed by the preprocessor
for:
- the simulation of backward models with “simul_backward”
- the perfect foresight simulation of purely backward/forward/static models

Also note that in this case, the preprocessor now defaults to “mfs=3” (i.e. it
minimizes the set of feedback variables and tries to renormalize equations).

This replaces the previous algorithm based on Dulmage-Mendelsohn (dmperm), plus
an ad hoc identification of some equations that can be evaluated (those with a
LHS equal to a variable, the log of a variable, or the diff-log of a variable).

By the way, the block_trust_region MEX has been modified so that it accepts a
boolean argument to decide whether it performs a Dulmage-Mendelsohn
decomposition (if not, then it performs a simple trust region on the whole
nonlinear system).

This provides a significant performance improvement (of almost an order of
magnitude for solve_algo=14 on a 700 equations model).

If the function to be solved returns a sparse Jacobian, simply convert it to a
dense representation.

sim1.m: use operation that preserves sparsity

See merge request !2111

Ref. !2019

Ref. !2110

Was crashing if the auxiliary model contains a constant.

See https://forum.dynare.org/t/iterative-ols-for-pac-equation/21379.

Fix: scale the matrices in tests/riccatiupdate.m to control the size of errors

See merge request !2108

Ref. !2105

Ref. !2103

Also capitalize MATLAB.

[skip ci]

[skip ci]

[skip ci]