The default epsilon limit (1e-5) may be changed by declaring a value for
FIT_LIMIT
When the sum of squared residuals changes between two iteration steps by a factor less than this number (epsilon), the fit is considered to have 'converged'.
The maximum number of iterations may be limited by declaring a value for
FIT_MAXITER
A value of 0 (or not defining it at all) means that there is no limit.
If you need even more control about the algorithm, and know the Marquardt-Levenberg algorithm well, there are some more variables to influence it. The startup value of `lambda` is normally calculated automatically from the ML-matrix, but if you want to, you may provide your own one with
FIT_START_LAMBDA
Specifying FIT_START_LAMBDA as zero or less will re-enable the automatic selection. The variable
FIT_LAMBDA_FACTOR
gives the factor by which `lambda` is increased or decreased whenever the chi-squared target function increased or decreased significantly. Setting FIT_LAMBDA_FACTOR to zero re-enables the default factor of 10.0.
Other variables with the FIT_ prefix may be added to fit, so it is safer not to use that prefix for user-defined variables.
The variables FIT_SKIP and FIT_INDEX were used by earlier releases of `gnuplot` with a 'fit' patch called `gnufit` and are no longer available. The datafile every modifier provides the functionality of FIT_SKIP. FIT_INDEX was used for multi-branch fitting, but multi-branch fitting of one independent variable is now done as a pseudo-3D fit in which the second independent variable and using are used to specify the branch. See multi-branch.