Hi David,It wasn't until I ran the Richard's steady state solution inside the docker container that I got output into my local mounted directory. When I ran the mpirun script with the docker launch script, it completed but didn't provide any output other than screen output. The issue with the python notebook is that it requires an ATS_SRC_DIR to import the ATS file reader tools, like ats_xmdf. I tried to get around this by just pulling the amanzi and ats git repos into a local folder without building them. There seems to be a link between the amanzi/src/physics/ats folders that will likely only exist when they are both built though.If you have suggestions for getting around this, or if I'm using the python notebook incorrectly, please let me know.Thanks,LizOn Tue, May 18, 2021 at 1:58 PM David Moulton <moultond...@gmail.com> wrote:Hi Liz,The way we set up the Docker image is so you can use it as if it were a local executable on your system. This means we don't intend you to be working in the image (which does not have matplotlib), but rather work on your local system in the python/jupyter environment there, and use the docker image to execute the ATS runs. It's true we've tested and worked with more on the command line so I probably need to provide much better instructions for both command line and jupyter.I hope this makes sense. Let me know if you have questions and I'll work on updating the wiki page on using Docker images.Thanks, David
--
You received this message because you are subscribed to the Google Groups "Amanzi-ATS Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ats-users+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ats-users/7bad30af-4043-4bf4-81cd-d6c5d2b42a56n%40googlegroups.com.