EDM#

class zfit.minimize.EDM(tol, loss, params, name='edm')[source]#

Bases: ConvergenceCriterion

Estimated distance to minimum.

This criterion estimates the distance to the minimum by using

\[EDM = g^T \cdot H^{-1} \cdot g\]

with H the hessian matrix (approximation) and g the gradient.

This is the same criterion as iminuit uses internally as well.

Parameters:
  • tol (float) – Tolerance for the criterion. If the criterion value is below the tol (usually), it is converged.

  • loss (ZfitLoss) – loss that will we minimized.

  • params (TypeVar(ParamTypeInput, zfit.core.interfaces.ZfitParameter, Union[int, float, complex, Tensor, zfit.core.interfaces.ZfitParameter])) – Parameters that will be minimized.

  • name (str | None) – Human readable name or description.

calculate(result)#

Evaluate the convergence criterion and store it in last_value

Parameters:

() (result)

converged(result)#

Calculate the criterion and check if it is below the tolerance.

Parameters:

result – Return the result which contains all the information

Returns: