On 2/23/2005 a NEURON user wrote:
> Is there any way to use batch_run with a variable time step?
On 2/23/2005 Michael Hines replied:
if cvode is active it is the integrator for batch_run. However it still breaks at
tstep step size intervals. A more direct integration that avoids fadvance is
cvode.solve(tstop)
which returns when tstop is reached. If you record values using Vector.record
then they will be recorded at the cvode chosen variable time steps.
(so be sure to tvec.record(&t) as well).