I have problem about the irfs.

[azuo1001]

I have problem about the irfs.
I have a bond demand function,
and its loglinerize form function is lambd=Et (lambd(+1)+Rb_bar*Rb(+1)/(1+Rb_bar));
Can I from ,get lambd(-1)=lambd+Rb_bar*Rb/(1+Rb_bar) ?
If I can, it means that Rb is determined by the change in ,since lambd(-1) is already determined in period t-1. if the above is right, Rb=- lambd*(1+Rb_bar)/ Rb_bar.
In my simulation, I got the first period irf of lambd was -1.987696853670524e-04 from shock to e_g,and the first period irf of Rb was -51.012917642318740, violating the function Rb=- lambd*(1+Rb_bar)/ Rb_bar, from which the first period irf of Rb should be -0.0137.

Anyone can help me, thank you very much!