README

This is the README file of the s2Dcd, a Python package that allows to obtain 3D multiple-point statistics (MPS) simulations by using only 2D training images (TIs). More details about the methodology can be found in the paper by Comunian, Renard and Straubhaar, DOI: 10.1016/j.cageo.2011.07.009.

Requirements

This version of s2Dcd heavily relies on the geone module. You can find more info about it and installation instructions at the link https://github.com/randlab/geone.

Quick start installation

NOTE: To run the s2Dcd, a MPS simulation engine like the DeeSse is needed! If you have installed the module geone then that should be already included. Note however that the version of the DeeSse included in the geone repository has some limitations. If you want to use the full functionalities of the package, please ask for a license.

Clone or download the s2Dcd package on a local directory, by using for example:

git clone [email protected]:randlab/s2Dcd.git

or

git clone [email protected]:alecomunian/s2dcd.git

(actually, the two repositories should contain exactly the same version of the code.)

Then inside the downloaded directory

pip install .

If the installation worked properly, then you should be able to perform an import s2Dcd from a Python console without any error/warning.

Examples

Animation

For an animation that illustrates how the s2Dcd approach works, check this link.

Simple example

Have a look at examples/01_Strebelle/s2Dcd_run-ex01.ipynb for a commented Jupyter notebook. You can also find the same file as Python script in examples/01_Strebelle/s2Dcd_run-ex01.py.

More info

Maintainers

At the moment, the code is maintained by A.Comunian. Don't hesitate to contact him if you have some suggestions of questions about the s2Dcd.

Source

The source file of the s2Dcd package is available both on GitHub and Bitbucket at the following links:

• https://github.com/randlab/s2Dcd
• https://bitbucket.org/alecomunian/s2dcd