How to handle warnings in SimBiology within series of simulations?

1 view (last 30 days)
To perform a sensitivity analysis I am running *.sbproj coded models using 'biosimulate' in a for-loop and see for some parameter combinations the following warning:
Warning: The right-hand side of the system of SimBiology ODEs results in complex numbers. The imaginary part of the result will be ignored.
> In sbioodeflux_298d036b_ba03_4469_9210_25b32c7bbb2b (line 44)
In sbiosimulate (line 140)
In GSA (line 107)
Can I trust the solutions? If not, how can I exclude the solutiuons which resulted in the warning above, i.e. is there a way to find programatically if a warning was produced?
Update
I added assignment of NaN's in case of fatal errors but still don't know how to handle the warnings.
try
...
[t,y] = sbiosimulate(m1);
f1(i,j) = trapz(t,y(:,varNo)); % f1 AUC
f2(i,j) = y(end,varNo); % f2 endPoint
catch me
disp( getReport( me, 'extended', 'hyperlinks', 'on' ) )
f1(i,j) = NaN;
f2(i,j) = NaN;
disp('Found an error-----------------------')
end

Accepted Answer

Arthur Goldsipe
Arthur Goldsipe on 3 Sep 2019
Hi,
It's difficult to say for sure whether you can trust the simulation results when you see this warning. I would definitely encourage you to try to understand why you are getting complex numbers. I see this most commonly when a model using exponentiation (such as with Hill kinetics) and a concentration becomes negative. To prevent this warning from occurring, I will sometimes replace terms like x^n with max(0,x)^n.
If you decide you need to programmatically detect when a simulation issues a warning, you can use the lastwarn function. For example, here is how I would modify your sample code to turn any warning that occurs during simulation into an error:
try
%...
lastwarn('', '');
[t,y] = sbiosimulate(m1);
[msg, msgID] = lastwarn;
if ~isempty(msgID) % or check for a specific warning ID
error(msgID, msg);
end
f1(i,j) = trapz(t,y(:,varNo)); % f1 AUC
f2(i,j) = y(end,varNo); % f2 endPoint
catch me
disp( getReport( me, 'extended', 'hyperlinks', 'on' ) )
f1(i,j) = NaN;
f2(i,j) = NaN;
disp('Found an error-----------------------')
end

More Answers (0)

Communities

More Answers in the  SimBiology Community

Categories

Find more on Perform Sensitivity Analysis in Help Center and File Exchange

Products

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!