GEONE is a python package providing a set of tools for geostatistical and multiple-point statistics modeling, comprising:
- multiple-point statistics (MPS) - DEESSE wrapper
- Gaussian random fields (GRF)
- other classical geostatistical tools (two-point statistics analysis (covariance, variogram, connectivity) / simulation (SGS, SIS) / estimation (kriging))
- pluri-Gaussian simulation
- other algorithms based on random processes (Poisson point process, Chentsov simulation)
git clone https://github.com/randlab/geone.git
cd geone
pip install .
Remove geone
pip uninstall -y geone
Note: first remove the directory 'geone.egg-info' from the current directory.
Do not launch python from the directory where the installation has been done (with pip
), otherwise import geone
will fail.
The following python packages are used by GEONE (tested on python 3.10.6):
- matplotlib
- multiprocessing (for parallel processes)
- numpy (tested with version 1.23.1)
- pyvista (tested with version 0.36.1)
- scipy (tested with version 1.9.1)
- GEONE includes a DEESSE wrapper to directly launch DEESSE within python. The DEESSE version provided with GEONE is a test version with restricted capabilities. To unlock the full capabilities of DEESSE, the user must obtain a commercial or academic license from the University of Neuchâtel. See LICENSE file for details.
- DEESSE and some other geostatistical tools provided by GEONE are compiled in C for windows and linux, and for python3.6 to python3.10. Note that for linux, the provided libraries depend on the library GLIBC 2.35, hence the library GLIBC of your OS has to be compatible with that version to ensure proper operation of GEONE.
Some modules in the package GEONE can be run as a script ('__main__' scope) and provide examples by this way.
Various examples are provided (notebooks in 'examples' directory) to get started with GEONE, as described below.
- Multiple-point statistics - simulation using the DEESSE wrapper:
ex_deesse_01_basics.ipynb
: basic DEESSE (categorical) simulationsex_deesse_02_additional_outputs_and_simulation_paths.ipynb
: retrieving additional output maps and setting the simulation pathex_deesse_03_search_neigbhorhood.ipynb
: advanced setting for the search neighborhood ellipsoidex_deesse_04_continous_sim.ipynb
: continous simulationsex_deesse_05_geom_transformation.ipynb
: simulations with geometrical transformations (rotation / scaling)ex_deesse_06_proba_constraint.ipynb
: simulations with probability (proportion) constraintsex_deesse_07_connectivity_data.ipynb
: simulations with connectivity dataex_deesse_08_multivariate_sim.ipynb
: bivariate simulations - stationary caseex_deesse_09_multivariate_sim2.ipynb
: bivariate simulations - setting an auxiliary variable to deal with non-stationarityex_deesse_10_incomplete_image.ipynb
: reconstruction of an image using a training data setex_deesse_11_using_mask.ipynb
: simulation using a maskex_deesse_12_multiple_TIs.ipynb
: simulation using multiple training imagesex_deesse_13_inequality_data.ipynb
: simulations with inequality dataex_deesse_14_rotation3D.ipynb
: simulations with rotation in 3D
- Elementary covariance models:
ex_elementary_cov_model.ipynb
: illustrations of elementary covariance/variogram models (in 1D)
- Gaussian random fields (GRF):
ex_grf_1d.ipynb
: example for the generation of 1D fieldsex_grf_2d.ipynb
: example for the generation of 2D fieldsex_grf_3d.ipynb
: example for the generation of 3D fields
- Variogram analysis tools and kriging:
ex_vario_analysis_data1D.ipynb
: example for variogram analysis and ordinary kriging for data in 1Dex_vario_analysis_data2D_1_omnidirectional.ipynb
: example for variogram analysis and ordinary kriging for data in 2D (omni-directional)ex_vario_analysis_data2D_2_general.ipynb
: example for variogram analysis and ordinary kriging for data in 2D (general)ex_vario_analysis_data3D_1_omnidirectional.ipynb
: example for variogram analysis and ordinary kriging for data in 3D (omni-directional)ex_vario_analysis_data3D_2_general.ipynb
: example for variogram analysis and ordinary kriging for data in 3D (general)
- Simulation and estimation with kriging (GeosClassic wrapper):
ex_geosclassic_1d.ipynb
:example in 1D for two-point statistics simulation and estimationex_geosclassic_1d_non_stat_cov.ipynb
:example in 1D with non-stationary covariance modelex_geosclassic_2d.ipynb
:example in 2D for two-point statistics simulation and estimationex_geosclassic_2d_non_stat_cov.ipynb
:example in 2D with non-stationary covariance modelex_geosclassic_3d.ipynb
:example in 3D for two-point statistics simulation and estimationex_geosclassic_3d_non_stat_cov.ipynb
:example in 3D with non-stationary covariance modelex_geosclassic_indicator_1d.ipynb
:example in 1D for two-point statistics simulation and estimation of indicator variablesex_geosclassic_indicator_2d.ipynb
:example in 2D for two-point statistics simulation and estimation of indicator variablesex_geosclassic_indicator_3d.ipynb
:example in 3D for two-point statistics simulation and estimation of indicator variablesex_geosclassic_image_analysis.ipynb
:example for two-point statistics analysis (covariance, variogram, connectivity, ...) of images (maps)
- Pluri-Gaussian simulations:
ex_pgs.ipynb
: example of pluri-Gaussian simulations in 1D, 2D and 3D (categorical, conditional or not), based on two latent Gaussian fields
- Ohter algorithms based on random processes:
ex_randProcess.ipynb
: example of Poisson point process, and Chentsov simulation in 1D, 2D and 3D