Add warning for variance decomposition at order=2

At order=2, the variance decomposition is still based on order=1. This
commit introduces a warning. See issue #278
(https://github.com/DynareTeam/dynare/issues/278). This information that
the variance decomposition is only triggered at periods=0 has been added
to the manual.

doc/dynare.texi

@@ -3161,7 +3161,7 @@ period(s). The periods must be strictly positive. Conditional variances are give
 decomposition provides the decomposition of the effects of shocks upon
 impact. The results are stored in
 @code{oo_.conditional_variance_decomposition}
-(@pxref{oo_.conditional_variance_decomposition}).
+(@pxref{oo_.conditional_variance_decomposition}). The variance decomposition is only conducted, if theoretical moments are requested, i.e. using the @code{periods=0}-option. Currently, variance decompositions are only implemented for @code{order=1}. In case of @code{order=2}, Dynare will thus not display the variance decomposition based on a second order approximation, but on a first order approximation. 
 
 @item pruning
 Discard higher order terms when iteratively computing simulations of
@@ -4230,7 +4230,7 @@ period 1, the conditional variance decomposition provides the
 decomposition of the effects of shocks upon impact. The results are
 stored in
 @code{oo_.PosteriorTheoreticalMoments.dsge.ConditionalVarianceDecomposition},
-but currently there is not output. Note that this option requires the
+but currently there is no output. Note that this option requires the
 option @code{moments_varendo} to be specified.
 
 @item filtered_vars

matlab/disp_th_moments.m

@@ -61,7 +61,11 @@ if ~options_.noprint %options_.nomoments == 0
     dyntable(title,headers,labels,z,lh,11,4);
     if M_.exo_nbr > 1 && size(stationary_vars, 1) > 0
         disp(' ')
-        title='VARIANCE DECOMPOSITION (in percent)';
+        if options_.order == 2
+            title='VARIANCE DECOMPOSITION (in percent), based on first order approximation';            
+        else
+            title='VARIANCE DECOMPOSITION (in percent)';
+        end
         if options_.hp_filter
             title = [title ' (HP filter, lambda = ' ...
                      num2str(options_.hp_filter) ')'];

matlab/display_conditional_variance_decomposition.m

@@ -49,8 +49,13 @@ StateSpaceModel.order_var = dr.order_var;
 conditional_decomposition_array = conditional_variance_decomposition(StateSpaceModel,Steps,SubsetOfVariables );
 
 if options_.noprint == 0
-    disp(' ')
+  if options_.order == 2
+    disp(' ')                
+    disp('CONDITIONAL VARIANCE DECOMPOSITION (in percent), based on first order approximation')
+  else
+    disp(' ')                
     disp('CONDITIONAL VARIANCE DECOMPOSITION (in percent)')
+  end
 end
 
 vardec_i = zeros(length(SubsetOfVariables),exo_nbr);

matlab/stoch_simul.m

@@ -142,6 +142,9 @@ if options_.nomoments == 0
     elseif options_.periods == 0
         % There is no code for theoretical moments at 3rd order
         if options_.order <= 2
+        if options_.order == 2
+            warning('You have requested a second order approximation, but variance decompositions currently only allow for first order. Displaying decompositions at order=1 instead.')
+        end
             disp_th_moments(oo_.dr,var_list);
         end
     else