simpeg.optimization.SteepestDescent.minimize#

SteepestDescent.minimize(evalFunction, x0)[source]#

Minimizes the function (evalFunction) starting at the location x0.

Parameters:
evalFunctioncallable()

The objective function to be minimized:

evalFunction(
    x: numpy.ndarray,
    return_g: bool,
    return_H: bool
) -> (
    float
    | tuple[float, numpy.ndarray]
    | tuple[float, LinearOperator]
    | tuple[float, numpy.ndarray, LinearOperator]
)

That will optionally return the gradient as a numpy.ndarray and the Hessian as any class that supports matrix vector multiplication using the * operator.

x0numpy.ndarray

Initial guess.

Returns:
x_minnumpy.ndarray

The last iterate of the optimization algorithm.