load intermediate results in SlowIRFs /MedianIRFs

[naivej]

Hi Tom,

I often find myself asking for IRFs to shock a,b,c but later want to add a shock d.
But computing SlowIRFs /MedianIRFs to shock d requires repeating the baseline simulation, which is the most time-consuming step in my experience (when "drop" is relatively large).
Even if someone computes IRFs to all shock in one go, it is often needed to compute IRFs to the same shocks with the opposite signs to show the asymmetry. Or the same shocks with different sizes.

A potential solution to this use case could be to load the baseline simulation computed previously.