Internal and external optimizers¶
A “fitter” or “optimizer” is the algorithm used to carry out the actual optimization. In funcFit, there are “internal” and “external” optimizers. The internal optimizers are those invoked by a call to the fit-method of the OneDFit object. External optimizers take a OneDFit object as an object and carry out the fitting without actually becoming a part of the object.
While, quite generally, internal optimizers are very convenient, it may be a little easier to exert control over the screws determining the details of the algorithm, if it needs to be tweaked to solve a specific problem.
Internal optimizers¶
Internal optimizers are invoked by OndeDFit’s fit method. Basically, an internal optimizer is a callable object, which carries out the optimization when called. A base class for internal optimizers is provided by funcFit.
External optimizers¶
An external fitter is a fitting algorithm not implemented as a part of funcFit’s OneDFit class. Further algorithms can thus be supported, which may give, e.g., stricter control over individual steps or internal settings.
A convenient way to pass data to the external fitters is the “funcFit data set” FuncFit data set.
