Updating modules for API reference

[apivarunas]

Updating ipmag, pmag, and pmagplotlib in Pmagpy-docs repo in order to keep consistency.

[Swanson-Hysell]

How does this issue differ from: #6 ?

Edit: oh, I see, it is because we manually bring over the .py files. Would be nice to find a way to link without needing to do it manually.

[apivarunas]

Exactly, as per your edit. I'm looking into how to get them linked.

[Swanson-Hysell]

Cool. For the moment, we can go ahead and copy them over (including rockmag.py) and note what PmagPy version in the commit.

[apivarunas]

Fair enough, I've done it that way right now. Looks like there should be a file sync workflow we can tell the Github machine to do, that will create PRs when the reference files in the main PmagPy repo get changed. I just need to get up to speed on secrets.

[Swanson-Hysell]

@apivarunas Do we have to do something special to get the API to regenerate? I just updated pmag.py in this docs repo to reflect the docstring improvements I have been making. However, I am not seeing them reflected in the docs even though the Github action to generate the book looks to have run.

[Swanson-Hysell]

It looks like the commit didn't trigger the gh-pages Github action just the the action in main which is likely the culprit.

[apivarunas]

Weird, I feel as if that Action has triggered under similar circumstances before.

[Swanson-Hysell]

I think by forcing a reload of the dependencies within requirements.txt we can move forward without having a local copy of pmagpy (preferable) and have it be pulled in from the latest pip release (and that could be what is happening here). And it makes the most sense to have this API reflect the latest pip release.

- name: Install dependencies
  run: |
    pip install -r requirements.txt

I can plan on a new pip release tomorrow to check this out after addressing a couple other issues.

[apivarunas]

Interesting. I just merged a commit that was adding in the SVEI notebook to the Template Notebook (as a first step to getting example notebooks here, instead of main repo). Looks like both actions ran, but the API Reference still doesn't seem to reflect those recent changes. Hmmm.

[apivarunas]

@Swanson-Hysell After some messing around, it seems to me as if the Jupyter Book does not in fact build off the modules placed within the PmagPy-docs repo. Instead, its keying off a pip install, it seems. When I crowbarred it into using the files within the repo itself (i.e., the most up-to-date thanks to our manual efforts), it started hanging errors based off imports at the top of the files.

[Swanson-Hysell]

I think that this behavior is a good thing. It means that we don't need to have the local files and can rely on it pulling in the data from a pip install. The key will be to have the action force a new pip install when it is run to regenerate the API docs.

[apivarunas]

I think that this behavior is a good thing. It means that we don't need to have the local files and can rely on it pulling in the data from a pip install. The key will be to have the action force a new pip install when it is run to regenerate the API docs.

Yes, this is preferable. I guess the local files were a red herring. Lol.

Right. I'll have to think about this. Actions are sort of new to me. But quite handy.