simpeg.electromagnetics.natural_source.Simulation2DElectricField.getJtJdiag#

Simulation2DElectricField.getJtJdiag(m, W=None, f=None)[source]#

Return the diagonal of ${\mathbf{J}}^{\mathbf{T}}\mathbf{J}$.

Where $\mathbf{d}$ are the data and $\mathbf{m}$ are the model parameters, the sensitivity matrix $\mathbf{J}$ is defined as:

$$\mathbf{J}={\displaystyle \frac{\partial \mathbf{d}}{\partial \mathbf{m}}}$$

This method returns the diagonals of ${\mathbf{J}}^{\mathbf{T}}\mathbf{J}$. When the W input argument is used to include a diagonal weighting matrix $\mathbf{W}$, this method returns the diagonal of ${\mathbf{W}}^{\mathbf{T}}{\mathbf{J}}^{\mathbf{T}}\mathbf{J}\mathbf{W}$.

Parameters:

m(n_param,) numpy.ndarray

The model parameters.

W(n_param, n_param) scipy.sparse.csr_matrix

A diagonal weighting matrix.

ffrequency_domain.fields.FieldsFDEM, optional

Fields solved for all sources.

Returns:

(n_param,) numpy.ndarray

The diagonals.