同问

当然可以。

If you are doing 2nd approximation and IRF analysis, the following matlab could help you solve your problem.
yy is the IRF for all endogenous variables. The following are the way Dynare doing IRF calculation.
by default, replic = 50 ; long =40; drop =100; Chol_S are the standard deviation of all shocks, diagnoal.

yy=0;
for j=1:replic   %draw two shocks
    ex1(:,i_exo_var)=randn(long+drop,nxs)*chol_S;
    ex2 = ex1;
    %add one standard deviation of two shocks to exogenous shock path, assuming totally 4 shocks
     ex2(drop+1,:)=ex2(drop+1,:)+[0.01 0.01 0 0];
    y1=simult_(temps,oo_.dr,ex1,iorder); % using dynare built-in function to simulate the whole model
    y2=simult_(temps,oo_.dr,ex2,iorder);
    yy=yy+(y2(:,M_.maximum_lag+drop+1:end)-y1(:,M_.maximum_lag+drop+1:end));
end
yy=yy/replic;




if output is the 1st varaibles in declaration order, the yy(1,:) is the IRF of output respond to two shocks.

帮忙第七来

同问，楼主解决这个问题了吗？