damping_misfit#

PySeismoSoil.helper_mkz_model.damping_misfit(param_without_Gmax: tuple[float, float, float], damping_data: ndarray) → float[source]#

Calculate the misfit given a set of MKZ parameters. Note that the values in param are actually the 10-based power of the actual MKZ parameters. Using the powers in the genetic algorithm searching turns out to work much better for this particular problem.

Parameters:

  • param_without_Gmax (tuple[float, float, float]) –

    MKZ model parameters, in the order specified below:

    gamma_ref, s, beta

  • damping_data (np.ndarray) – 2D numpy array with two columns (strain and damping value). Both columns need to use “1” as the unit, not percent.

Returns:

error – The mean absolute error between the true damping values and the predicted damping values at each strain level.

Return type:

float