simpeg.maps.LogisticSigmoidMap.inverse#

LogisticSigmoidMap.inverse(m)[source]#

Apply the inverse of the mapping to an array.

For the logistic sigmoid mapping $\mathbf{u}(\mathbf{m})$, the inverse mapping on a variable $\mathbf{x}$ is performed by taking the log-odds of elements, i.e.:

$$\mathbf{m}={\mathbf{u}}^{-1}(\mathbf{x})=logit(\mathbf{x})=\log \frac{\mathbf{x}}{1-\mathbf{x}}$$

or scaled and translated to interval (a,b): .. math:

\mathbf{m} = logit(\frac{(\mathbf{x} - a)}{b-a})

Parameters:

mnumpy.ndarray

A set of input values

Returns:

numpy.ndarray

the inverse mapping to the elements in m; which in this case is the log-odds function with scaled and shifted input.