The advanced tab contains settings to fine tune the nonlinear fitting algorithm.

Change the defaults with caution, as improper values may lead to sub-optimal deconvolution process.

Iteration stopping criteria for fitting

Max iteration count

Upper limit for number of iterations for nonlinear fitter. Default value is 300. Less than this value may cause non-optimal fit results.

Minimum RXSQ value

Lower limit for reduced Chi-squared, when nonlinear fitter should quit unconditionally. Default value is 0.05 (this is almost never reached).

Minimum gradient change

Lower limit for gradient value change. If the gradient of fitted parameters changes to less than this value between iterations, then iteration will be stopped. Default value is 1e-9 (should not be greater than 1e-6).

Minimum line search derivative change

Lower limit for derivative value change. If the derivative of fitted parameters changes to less than this value between line search iterations, then iteration will be stopped. Default value is 1e‑5.

Minimum line search RXSQ change

Lower limit for reduced Chi-squared value change. If the reduced Chi-squared changes to less than this value between line search iterations, then iteration will be stopped. Default value is 1e-6.

Copyright © 1998-2007 by HyperLabs Software Budapest, Hungary