Pass current timestep to affect! to modify integrator values with lookup table

Specific Domains Modelling & Simulations
1

I am trying to simulate patient response to a drug and am currently fitting a model to patient data. However the frequency and strength of the dose changes frequently, so I have the time and strength for each dose recorded in a dataframe. Is it possible to pass the current timestep t to affect! so that it can lookup the value in the table?

example_doses = DataFrame([[0,3,10,40], [50,88,88,12]], ["time", "dose"])

function condition(u, t, integrator)
    t in example_doses.time
end

function affect!(integrator)
    integrator.u[10] = example_doses.dose[findfirst(==(t), example_doses.time)]
end

cb = ContinuousCallback(condition, affect!)

Note the actual doses dataframe has > 1000 rows and this must be simulated for many patients, so as far as I can see storing the doses in the integrator parameters to be accessed via integrator.p as mentioned here is not practical.

2

it’s just integrator.t to get the current time.

Though note that if you’re going to do this you need to set tstop. You may want to use the PresetTimeCallback for this.

1 Like
3

Thanks Chris, this was exactly what I was looking for.