Potential tweaks for consideration

[shimwell]

Just watched your nice video 📹

Congrats on the smooth workflow 🎉 and great to see that the model and work is available to all 👍

One of the great things about putting this on GitHub is that we can all pitch in with random ideas 😉

I noticed a few things that might be of interested to you 👀

• consider openmc.stats.CylindricalIndependent 🔴 instead of openmc.stats.Box 🟥 as this may provide more efficient source [docs link](sampling https://docs.openmc.org/en/stable/pythonapi/generated/openmc.stats.CylindricalIndependent.html)

• consider making some materials from a chemical formula using .Material.add_elements_from_formula which is in the docs and accepts enrichment options

• This is not always considered good practice but perhaps for your use case it is ok. Consider allowing jupyter notebook to run without login by adding the jupyter password into the entry point command something like this might be worth trying CMD ["jupyter", "lab", "--port=8888", "--no-browser", "--ip=0.0.0.0", "--allow-root", "--NotebookApp.token=''", "--NotebookApp.password=''"]

• consider using openmc.Model() which exports all the xml files needed and returns the name of the statepoint file #LineSaver.

• consider making the github.com/openmsr/openmc_install_scripts repository pubilc as then the .gitmodules will work for all and those who are keen can build the dockerfile instead of downloading 🔗

• Consider building the Docker image with some github actions CI and potentially running the notebook in the CI as well. Would be great to see some nice green badges in the readme.md ✔️

[ebknudsen]

Thank you for the kind words and suggestions. 👍 All good ones to keep in mind for next version of this one, or similar videos.
On the specific point on install scripts, we're currently reviewing and testing with the intention of making them public. At present they're more complicated than they have to be.

[shimwell]

On the topic of simplifying the install we have made some nice progress recently by adding DAGMC to the Conda install for OpenMC

This attached picture shows how DAGMC is installed by compiling it with OpenMC on the left, a hybrid DAGMC conda and openmc compile in the center and on the right an OpenMC conda install that includes DAGMC.

This reduction in installation complexity was achieved by this pull request and we can certainly add more packages to the conda install to simplify the install further.

The DAGMC bundled with OpenMC conda install doesn't currently include Embree or Double Down but these could be included with a little development effort on the feedstock repo and associated repos.

[ebknudsen]

That's great. Thank's for pointing this out.
Specifically in terms of "home-use" MSRE double down and embree are a practical necessity, since otherwise the complexity of the model makes it very slow to run without.