Strategy#

A strategy helps a minimizer in some situations - such as when encountering NaNs - and hooks into the loss evaluation. They can be used to create a Minimizer with a specific strategy.

zfit.minimize.PushbackStrategy()

Pushback by adding nan_penalty * counter to the loss if NaNs are encountered.

zfit.minimize.DefaultStrategy

alias of PushbackStrategy

zfit.minimize.DefaultToyStrategy()

Same as PushbackStrategy, but does not raise an error on full failure, instead return an invalid FitResult.