Repository navigation
Example Folder Cleanup #351
Description
The examples folder is quite messy right now, with what seems to be a variety of both examples and benchmark-like cases. Some also contain their own run/analysis files, but others don't. A specific way I can see to clean this up and ensure a standard moving forward is removing the individual run scripts in each example folder in favor of a master script.
A rough sketch of how I see this working:
We make a master script RunExample.jl:
julia RunExample.jl "DIIID-like_ideal_example"Within the main function of this file, it then checks the string passed in:
- If it ends in "example" it basically does the existing
run_example.jlfew lines - If it ends in "scan" it does a generalized version of Daniel's
run_scan.jlscripts
This would avoid having to maintain a run file for each example we maintain, and also force all examples conform to be either a "example" or "scan" type (or additional types as needed).
This doesn't account for example analysis scripts. We could add a generalized version of that too if needed
Activity
jhalpern30 commentedon Oct 6, 2026
I am actually going to remove GPEC v2.0 from this - I think this should be revisited after @matt-pharr is done their refactoring, since the long term goal is to make GPEC more scriptable which will probably update how we do examples. But I do like the general idea of a more unified examples interface once the refactoring is done
cc @logan-nc
matt-pharr commentedon Oct 7, 2026
@jhalpern30 I have one more PR coming in soon and then we should be able to address this! I think that this deserves the v2.0.0 since the examples we want to look good for people trying to understand how to use the code.
logan-nc commentedon Oct 27, 2026
Not opening the master-script PR now.
examples/has 22 entries with 14 per-folder scripts (7run_example.jl, 4analyze_example.jl, 2run_scan.jl, 1run_imas_example.jl), and three open API PRs still rewrite the decks: #569 deletes keys from every example, #574 edits the regularization lines, #581 moveskinetic_fileinto[Equilibrium]. A unified runner opened tonight would need re-stamping after each. Matt claimed this after his refactor; suggest doing it immediately after #581 lands, with theexample/scansplit proposed here.matt-pharr commentedon Oct 27, 2026
@logan-nc @jhalpern30 this is scheduled in #602