A question re: setting initial conditions as last states of previous simulation

[zuxfoucault Wong]

Dear TVBer,

When setting initial conditions as the last states of the previous simulation, is it reasonable to extract the last states with the length of time steps corresponding to good_history_shape[0] (which is also related to sim.connectivity.horizon)? Or it could be done by replicating the same last state multiple time steps with the length corresponding to good_history_shape[0]?

I am also wondering why the horizon time steps are required in the setting of the initial condition? Any suggestions would be appreciated! Thank you!

Best,

FuTe

[WOODMAN Michael]

Hi

The requirement issues from the use of delay differential equations because the white matter connectivity induces time delays.  The sim.history.buffer and sim.current_state  have the right content.  You can copy them into the next simulation for a deterministic integrator.  For stochastic, the random state also needs copying.  There is a demo which has an example with the keyword ‘continuation’.  

Cheers

Marmaduke

On 26 Aug 2021, at 22:30, zuxfoucault Wong <zuxfo...@gmail.com> wrote:

Dear TVBer,

--
You received this message because you are subscribed to the Google Groups "TVB Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tvb-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tvb-users/e1679780-7460-41d0-b4ca-3d19e3542ca2n%40googlegroups.com.