Dear Matlab community, I have a question about simbiology app in MATLAB

I am trying to fit a three compartmental model in MATLAB. Can anyone kindly tell me how to put bounds (lower bound and upper bound) on estimated parameters while fitting the model ? Thanks a lot and really appreciate your help.

Answer 1

[Arthur Goldsipe]

The answer depends on what you mean by "a fit." If you are performing nonlinear regression, then you can specify bounds. See this post for more details.

However, bounds are not currently supported if you are performing nonlinear mixed effects modeling (for example, as implemented in the function sbiofitmixed in SimBiology or nlmefit and nlmefitsa in Statistics Toolbox).

-Arthur

[chevy]

Thanks a lot Arthur for the response. Actually, I am trying to perform non linear mixed effect modeling and using NLMEFIT (population) in simbiology app and want one of the parameters to stay in physiologically feasible range. Anyways, thanks again for the answer.

[Arthur Goldsipe]

Ah, ok. As a workaround, you could estimate an unbounded parameter and converted it to a bounded parameter using a repeated assignment. For example, you could estimate unbounded X and convert it to a bounded parameter Y with lower bound A and upper bound B by introducing the following repeated assignment:

Y = A + (atan(X)/pi + 0.5)*(B-A)