We're wanting to run a disease model where we start the model in the steady state. It would be too difficult to solve for the steady state algebraically, however, and we would prefer to not start the model 100 years or so earlier to find it because some of the parameters are sensitive to time so this may skew later results. Is there a way in vensim to simulate for the initial condition values to start in the steady state?
Thanks!
Heather
steady state calculations
-
bob@vensim.com
- Senior Member
- Posts: 1107
- Joined: Wed Mar 12, 2003 2:46 pm
The SIMULTANEOUS and FIND ZERO functions do not operate on initial computations but only for active computations - so unfortunately these will not work for what you are trying to do.
One thing you can do, however, would be to simulate the model for a while to get to a steady state, then resume from that simulation for experimental runs. This is a bit cumbersome, but should work for a stable system. An alternative approach is to solve out the algebra with a bit of tweaking here and there. Although this may sound overwhelming, if you are clever about parameterizations it is often possible to arrive at relatively straightforward solutions.
One thing you can do, however, would be to simulate the model for a while to get to a steady state, then resume from that simulation for experimental runs. This is a bit cumbersome, but should work for a stable system. An alternative approach is to solve out the algebra with a bit of tweaking here and there. Although this may sound overwhelming, if you are clever about parameterizations it is often possible to arrive at relatively straightforward solutions.