Update loading documentation for user feedback

[ascourtas]

As a User, I need good documentation to know what's going on, especially after a major code refactor that will break all of my stuff

We've made a lot of changes in the Refactor, but we need to update the documentation so users can actually get using it!

Assumptions:

1. We're just doing this for loading, not publishing atm until the publishing changes are merged into dev
2. This will go out in an alpha release with the loading changes, but currently those are all in dev
3. We're still using GitBook
4. Related to #410

Acceptance Criteria

1. Given the documentation, a close collaborator (eg UW) can
2. The documentation accurately reflects the current state of the dev codebase

Tasks

• [x] update all GitBook docs related to loading/downloading datasets
• [x] double check example notebooks are correct (see Investigation below)
  • [ ] update notebooks accordingly
• [ ] ensure that GitBook descriptions of metadata, data schemas, etc match the new formats
• [x] generate class- and method-level documentation based on docstrings

[ascourtas]

Example notebook investigation

Notebooks that need amending:

• [ ] PACBEDCNN-thickness-mistilt

  • the original version of this notebook was mostly used to demonstrate model functionality; however, in the newer version of this notebook, the model functionality has been removed since we removed Foundry models (makes sense)
  • this makes the notebook fairly meaningless
  • Proposed solution: the original version should be ported over to Garden to be used as a seedling

• [ ] atom-position-finding

  • need to update language about Globus account, "needing" GCP, etc (since HTTPS is the default download protocol now, and users don't need to install GCP)
  • add a little text blurb at the bottom tying everything together

• [ ] bandgap

  • delete excessive cell output
  • update language about Globus account and GCP (same as atom-position-finding)
  • add closing sentence
  • might want to replace search() approach with get_dataset_by_name() or DOI (and maybe still mention search, so the user knows what's up)

• [ ] dendrite-segmentation

  • same as bandgap

• [ ] g4mp2-solvation

  • same as bandgap

• [ ] oqmd

  • same as bandgap
  • also, there's an error loading the dataset? the error doesn't make much sense, but I'm guessing it might be because the default is to load by HTTPS and the oqmd dataset is too large
  • Proposed solution: try loading the dataset again, and if that doesn't work switch from https to globus

• [ ] qmc-ml

  • same as bandgap
  • also some weirdness about a warning when doing the installs? need to look into it

• [ ] zeolite

  • update language about Globus account and GCP (same as atom-position-finding)
  • add closing sentence

publishing-guides aren't in scope for this issue

[kjschmidt913]

Gitbooks is awful and has been difficult to update for a dev release.