The following excerpt from imcforecast.m seems to be causing a crash with conditional_forecast:
[quote]for i = 1:n1
idx = [idx ; oo_.dr.inv_order_var(strmatch(deblank(constrained_vars(i,:)),M_.endo_names,‘exact’))];
jdx = [jdx ; strmatch(deblank(options_cond_fcst.controlled_varexo(i,:)),M_.exo_names,‘exact’)];
mv = zeros(n1,NumberOfStates);
mu = zeros(ExoSize,n2);
mv(i,idx(i)) = 1;
mu(jdx(i),i) = 1;
I think the problem is with
[quote] idx = [idx ; oo_.dr.inv_order_var(strmatch(deblank(constrained_vars(i,:)),M_.endo_names,‘exact’))];
I bypass this problem by directly inserting the constrained_vars somewhere and it works. Also there seems to be a problem with plot_icforecast.m
What I did was to save the output before calling this function, but it’s not convenient.
Can one carry out k-step ahead analysis here too and how?