r2ogs5 is an R-API to the opens-ource multiphysics simulator OpenGeoSys 5. r2ogs5 allows to

• reading and preparing input files
• generate simple structured meshes
• executing single and ensemble simulation of serial and parallelized OGS models
• retrieving, post-processing and visualizing OGS output data
• calibrate OGS simulation models with Bayesian Optimization

Package installation was tested for Linux (Ubuntu 20.4) and Windows, package functionalites are only tested for Linux operating systems so far.

Before you install and load r2ogs5 you need to set-up your python environment. This is required to read in the .vtu and .pvd files produced by OpenGeoSys 5. You need a Python installation including the libraries numpy and vtk. If you have an Anaconda installation on your system, setting up a Python environment with numpy and vtk will work. r2ogs5 was tested with the following environment:

$ conda create -n r2ogs5 python=3.8
$ conda install -n r2ogs5 -c conda-forge vtk numpy
$ conda activate r2ogs5

The configuration of avialable python environments on the system can be found via

reticulate::conda_list()

In R, the path for the python environment needs to be indicated to the package reticulate. This can be achieved e.g. via:

Sys.setenv(RETICULATE_PYTHON = "C:/Users/admin/anaconda3/envs/r2ogs5/python.exe")

If you don't have any Python installation on your system, you can download the latest release of Python from here. Instructions on how to install numpy and vtk can be found here and here.

The first option for installation is directly from the online repository via

devtools::install_git("https://gitlab.opengeosys.org/ogs5/r2ogs5")

or alternatively

remotes::install_git("https://gitlab.opengeosys.org/ogs5/r2ogs5")

in your R console.

The second option is first cloning the repository by typing in a terminal

git clone https://gitlab.opengeosys.org/ogs5/r2ogs5.git
cd r2ogs5
R --slave -e 'devtools::install(".")'

Note that it is necessary to have git https://www.git-scm.com/ and/or the R-package devtools installed on your system.

Finally, the actual simulation program, OpenGeoSys 5 needs to be available on the system. r2ogs5 does not come with OGS5 executables, as they usually need to be compiled specifically for the machine run on. For this tutorial, the default compiled executables should be sufficient which can be downloaded from here for Linux, macOS, and Windows.
The name (here ogs_fem), and location of the OGS5 executable have to be set accordingly via options.

options(r2ogs5.default_ogs5_bin = "path/to/ogs_fem")

Note: To install r2ogs5 including vignettes, the packages knitr, rmarkdown with pandoc need to be installed on the system. In addition, an ogs executable named "ogs_fem" needs to be provided in the search path of the OS e.g. in /usr/local/sbin, /usr/local/bin /usr/sbin, /usr/bin. This is currently only possible for Linux operating systems.

For using the package please consider the vignettes:

• base package functionality
• define ensemble runs
• model calibration with Bayesian Optimization

R scripts showing you how to recreate the official OpenGeoSys 5 benchmarks with r2ogs5 are available here. These scripts can be used in a similar fashion as desribed in base package functionality.

See this repository for the model input files and R-scripts used to obtain the results of the example in the articel published in Groundwater.

• please report bugs via opening an issue
• or ask question on our OGS Q&A page

If you want to contribute to r2ogs5 look here, please.