Hi Abed,
when evaluating expressions of observables, SimBiology first tries to determine whether the result is a scalar value or a vector. I therefore recommend to write observables' expressions in a way that they either always return a scalar value or always a vector of values - for any possible input values of referenced model components and time.
In your case, the expression max(A(time > 3*7*24))) can be a vector whose length differs from the length of the time vector, e.g. if none of the values in time exceeds 3*7*24, or if the maximum is attained at multiple time points. You can write a helper function that returns the max. value and protects against invalid edge cases:
function maxValue = getMaxValueAfterSpecifiedTime(values, time, threshold)
% Get max value (call unique to ensure scalar or empty output).
maxValue = unique(max(values(time > threshold)));
% Protect against empty maxValue by returning nan.
if isempty(maxValue)
maxValue = nan;
end
end
You can then call this helper as follows:
% Load/build your model
mdl = sbmlimport("radiodecay");
% Observable referencing time point
observable = addobservable(mdl, "maxX", "getMaxValueAfterSpecifiedTime(x, time, 2)");
set(observable, "Active", true);
% Set simulation options
configset = getconfigset(mdl);
configset.RuntimeOptions.StatesToLog = "x";
% Simulate and plot model state and active observable
simData = sbiosimulate(mdl);
sbioplot(simData);
When you run this particular example, you may see that maxX differs from the actual max value as shown in the plots. You can add interpolation to your helper function to increase the accuracy of the reported max value; an example how this could look like is attached to this post.
I hope this helps.
Best,
-Florian
