
Optimizer parameters
This page describes the optimization parameters.
- Standard error
- Optimization stops when dispersion of values falls below this value. Dispersion is defined as sum (f - mean(f))^2 For the Levenberg-Marquardt method, f is the set of function value for the last 10 iterations. For the Nelder-Mead method, f is the set of function values at the simplex vertices.
- Step size
- optimization stops when step size falls below specified value
- Function value
- optimization stops when function value falls below specified value
- Iteration limit
- optimization is aborted if more iterations that this
- Universe radius
- length of vector formed from all variable parameters must remain less than this