because my use of the term “reverse” is not in compliance with any literature I will close this portion agreeing that it is a mistake calling it as such
yes, and not log-scaling everything worked in this particular example
sorry,
$fct_rn
is 1E-3 a preselected scaling factor to bring the optimization target closer to unity, and $k_deg
is reaction-rate constant estimated experimentally , everything else is either a compound (cat_a, cat_b, H2A, DHA, H2
) or a reaction-rate constant to be optimized (k_1, k_2, k_4
)
oh, yes definitely, I am doing this quite a lot, to the point wondering how come I hadn’t stumbled on it earlier.