Run Model One Time Step

[akuptel]

Hi,
Very simple question and my apologies if this isn't the right forum, but I wanted to ask how to run a model a single time step (in order to implement an RL algorithm). I saw that .run() has arguments for final_time and initial_conditions, but I was confused how to specify the initial time using the latter.

Thanks!

Update: I tried just repeatedly using _euler_step(dt = TIMESTEP), but when I do that my model just stays in the initial conditions

[akuptel]

I managed to figure out a workaround, but I'm not sure I'm entirely pleased with it-- I just iterated through each of the time steps, at each time beginning at the "current" configuration, set a parameter equal to a custom function at each time step (for some reason I can't set it equal to it at the beginning of the loop). (I'm using the SEIRb model for this.)

I suppose I wish there were a way to store the current state of the model of every variable or something, it just seems a bit odd to just use initial_conditions = 'current.'

[enekomartinmartinez]

Hi @akuptel

Great you found out how to do that.
Yes, the way to run only one time step is to set the final_time = current_time + time_step and use the "current" or "c" initial_condition option to keep the values of the model from one iteration to the next.

Let us know if we can help you with anything else :)

[rogersamso]

Desar @akuptel release 3.11 of PySD should be the solution to your question.

Please check this section or the docs for details.

Best