This is the source code to the static site visualizer for the 11th SPE Comparative Solutions Project. For more details see the website of the study..

The source code is useful if you want to see how the the visualizations in the page were made, or if you want to make a version with your own simulation results included.

Example from spe11b/OPM4

Building this yourself requires installation of Julia. Note that if you are building on a remote/server machine you may have to use xvfb-run julia to avoid OpenGL errors as some of the underlying packages assume that a display is available.

Run julia from the docs folder:

cd docs/
julia --project=.

Develop the helper package, which should also install all other dependencies needed to build the project.

]dev ../CSP11Visualizer

Put the data in a folder under the root folder of the repository named data, or place the absolute path to the data folder in a textfile data/path.txt. The complete dataset of all submissions can be downloaded from DOI 10.18419/DARUS-4750.

The resulting folder structure should look like this, taking IFPEN result 2 for case b as an example: data/spe11b/ifpen2/

There are two parts to getting the website built. First, you should generate the animations, and then you can build the entire website. Please note that the movies take a long time to build for all results and all participants.

include("generate_movies.jl")
include("build.jl")

cd docs/
julia --project=.

and then

include("build.jl")

The resulting files will be in the docs/build folder and can be opened in a browser.

Building the website for all the results in the benchmark takes a long time. The biggest culprit is the generation of movies, which is for this reason a separate build step. Here are some hints if you want to speed up build times.

• You can optionally skip the generate_movies - the website will still be be functional, but missing the movies.
• Both generate_movies and build can be edited to only publish specific cases. In either case this is done by setting build_all_dense=false at the top of the script and editing the cases_* variables to include the cases you want. If you just want to quickly test the build, the default configuration for build_all_dense=false is to build one participant result for each case (A/B/C).
• All pages that make plots are actually just templated scripts that can be run manually. They are found in scripts, together with some other experimental scripts that may or may not be useful. The built pages use sparse_*_static.jl and dense_*_template.jl. Building the actual webpages pages are reliant on finding and replacing groupname = "opm" and resultid = 1 from these scripts, so manually editing these before building the page is not recommended.