Skip to content

[R-sig-dyn-mod] iterative procedure vs deSolve

1 message · Paula Sanchez Marin

#
Thank you Johannes, I've tried your suggestion of changing rtol and atol 
values but the deSolve solution is always the same.

In fact, I have obtained the answer to my question from Thomas Petzoldt 
and Stephen P. Ellner (thank you!).

I copy here the answer from Stephen:

"the problem is that this
is not the analytic solution to the differential equations. The first line is the solution to the first differential equation, assuming that
Crest is constant. The second line is the solution to the second differential equation, assuming that Cgills is constant. But since
Cgills and Crest change over time, those are not the solution to the system of differential equations. The answer to trust is the
one that you're getting from the ODE solvers in R."