Skip to content


This package provides methods for inverse analysis using parameter fields that are represented using geostatistical (stochastic) methods. Currently, two geostatistical methods are implemented. One is the Principal Component Geostatistical Approach (PCGA) proposed by Kitanidis & Lee. The other utilizes a Randomized Geostatistical Approach (RGA) that builds on PCGA.

Randomized Geostatistical Approach (RGA) references:

Two versions of PCGA are implemented in this package

  • pcgadirect, which uses full matrices and direct solvers during iterations
  • pcgalsqr, which uses low rank representations of the matrices combined with iterative solvers during iterations

The RGA method, rga, can use either of these approaches using the keyword argument. That is, by doing rga(...; pcgafunc=GeostatInversion.pcgadirect) or rga(...; pcgafunc=GeostatInversion.pcgalsqr).

GeostatInversion.jl module functions:

```@autodocs Modules = [GeostatInversion] Order = [:function, :macro, :type]

<a id='Module-GeostatInversion.FDDerivatives-1'></a>

## Module GeostatInversion.FDDerivatives

GeostatInversion.FDDerivatives module functions:

Modules = [GeostatInversion.FDDerivatives]
Order   = [:function, :macro, :type]

Module GeostatInversion.RandMatFact

GeostatInversion.RandMatFact module functions:

```@autodocs Modules = [GeostatInversion.RandMatFact] Order = [:function, :macro, :type]

<a id='Module-GeostatInversion.FFTRF-1'></a>

## Module GeostatInversion.FFTRF

GeostatInversion.FFTRF module functions:

Modules = [GeostatInversion.FFTRF]
Order   = [:function, :macro, :type]