Closed steven-murray closed 1 year ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
@jsdillon this is passing tests. The only comment that may be unresolved is the one about phs_sol
, which I can't think of a better way to do.
This adds a small test file and runs testing on the file_calibration notebook. We can add other notebooks as well with a similar framework, if deemed necessary.
There are two downsides to the current setup, as I see it:
vis_cpu
(for which we need GPU's). I can set that up if we think that's a better way.git pull
the repo and re-run the updated notebooks without installing the package itself. This means that any updated dependencies that might be reflected in thepyproject.toml
in the lastest commit would not be installed in the user's env, which can cause the notebooks to break (especially with the rapid and entwined development going on in hera_qm and hera_cal). That is, if you update the notebook to do something cool and new, and tell me to update, then I go do agit pull
and re-run my makeflow that points to the local copy of the repo, it will get the updated notebook, but not the updated dependencies. The better way to do this, IMHO, would be to include thenotebooks/
directory inside the package itself, and include a CLI script (or scripts) that get installed that execute the notebooks. Then the common usage when you want to update and run the notebook would be to install the newest version of the package from github withpip
, and then run the script, eg.heranb run file_calibration
.Doing both of these things would be a much more robust way of ensuring that this test actually tests what we'll be running (either on site or at NRAO)