MPSLIB: A C++ class for sequential simulation of multiple-point statistical models. Geostatistical simulation methods allow simulation of spatial structures and patterns based on a choice of statistical model. In the last few decades multiple-point statistics (MPS) has been developed that allows inferring the statistical model from a training image. This allows for a simpler quantification of the statistical model, and simulation of more realistic (Earth) structures. A number of different algorithms for MPS based simulation have been proposed, each associated with a unique set of pros or cons. MPSLIB is a C++ class that provides a framework for implementing most of the currently proposed multiple-point simulation methods based on sequential simulation. A number of the most widely used methods are provided as an example. The single normal equation simulation (SNESIM) method is implemented using both a tree and a list structure. A new generalized ENESIM (GENESIM) algorithm is proposed that can act as (in one extreme) the ENESIM algorithm, and (in another extreme) similar to the direct sampling algorithm. MPSLIB aims to be easy to compile on most platforms (standard C++11 is the only requirement) and is released under the Open Source LGPLv3 License to encourage reuse and further development.
Keywords for this software
References in zbMATH (referenced in 3 articles , 1 standard article )
Showing results 1 to 3 of 3.
- Hansen, Thomas Mejer: Entropy and information content of geostatistical models (2021)
- Holm-Jensen, Tue; Hansen, Thomas Mejer: Linear waveform tomography inversion using machine learning algorithms (2020)
- Hansen, TM; Vu, LT; Bach, T.: MPSLIB: A C++ class for sequential simulation of multiple-point statistical models (2016) not zbMATH