simpeg.electromagnetics.natural_source.receivers.ApparentConductivity.evalDeriv#

ApparentConductivity.evalDeriv(src, mesh, f, du_dm_v=None, v=None, adjoint=False)[source]#

Derivative of data with respect to the fields.

Let $\mathbf{d}$ represent the data corresponding the receiver object. And let $\mathbf{u}$ represent the discrete numerical solution of the fields on the mesh. Where $\mathbf{P}$ is a projection function that maps from the fields to the data, i.e.:

$$\mathbf{d}=\mathbf{P}(\mathbf{u})$$

this method computes and returns the derivative:

$${\displaystyle \frac{\partial \mathbf{d}}{\partial \mathbf{u}}}={\displaystyle \frac{\partial [\mathbf{P}(\mathbf{u})]}{\partial \mathbf{u}}}$$

Parameters:

srcfrequency_domain.sources.BaseFDEMSrc

The NSEM source.

meshdiscretize.TensorMesh

Mesh on which the discretize solution is obtained.

fsimpeg.electromagnetics.frequency_domain.fields.FieldsFDEM

NSEM fields object for the source.

du_dm_vNone, optional

Supply pre-computed derivative?

vnumpy.ndarray, optional

Vector of size

adjointbool, optional

Whether to compute the ajoint operation.

Returns:

numpy.ndarray

Calculated derivative (n_data,) if adjoint is False, and (n_param, 2) if adjoint is True, for both polarizations.