Estimate model parameter values from observed data


fit <model parameter 1> <model parameter 2> …


The FIT command estimates the values of model parameters using data sets specified using the DATA command by minimizing the sum of the squared errors, where the errors are summed across observed data points and available data sets (where corresponding descriptor variables are taken into account).  Minimization is performed using the BOBYQA (Bound Optimization BY Quadratic Approximation) algorithm.  Optional upper and lower bounds are imposed on parameters using the SET command @min and @max flags.  Upon successful completion of parameter estimation, the optimized parameter values will be displayed in the command output window.


data @file='observed.csv' dataset1 t='T' x='X' tstop=60 c=12.0

set k1 = 0.2 @min=0.1 @max=0.35

set k2 = 0.4 @min=0.23 @max=0.45

set k3 = 0.05 @min=0.03 @max=0.09

fit k1 k2 k3