The workaround consists in appending the names of those macro-directives with a
regular space followed by a zero-width space (U+200B).

An extra LaTeX declaration was also added in order to make this hack work with
the PDF.

Ref. #1707

Fix pruned state space at third order

See merge request !1711

Related to #1706

✅ Add unit test for pruned state space up to order 3


//


Changed tolerance in new test 

... Which would probably need to be updated ;-)

geweke_moments.m: correctly use parameter instead of hard-coded value

See merge request !1710

[skip ci]

[skip ci]

Implicit expansion (a.k.a. automatic broadcasting) was introduced in MATLAB
R2016b (and it has been present in Octave for quite some time).

Hence use bsxfun() instead.

The problem had been introduced in 228b2a53.

Under Octave, the default seed in fs2000ns_uncorr_ME.mod leads to a generated
dataset that makes fs2000ns_corr_ME.mod and fs2000ns_corr_ME_missing.mod
fail (due to a large difference between univariate and multivariate diffuse
filters).

Factorization of discretionary policy

Closes #1705

See merge request !1709

conditional forecasts: add check whether simulation was successful

See merge request !1708

Related to Dynare/dynare#1173

Closes #1705

The steady state is always zero for discretionary policy. And, in the case of a
steady state file, this call would not be able to update parameters (since it
does not modify M_), nor would it need to do so (since this has already be done
earlier in the function).

Ref. #1705

Under GNU/Linux and macOS, double-quote arguments before passing them to the
shell. In particular, this allows passing single-quotes within those arguments.

We therefore have to escape the four characters that are interpreted within
double-quoted strings in POSIX shells: \, ", $ and `

On Windows, also systematically escape the backslashes.

Also move display of arguments before escaping, so that it remains readable.

Ref. #1696

[skip ci]

Fixes related to estimation under discretion

See merge request !1707

imcforecast.m: make sure that draw index for CIs stays within feasible bounds

See merge request !1706