develop more guidance on how to fill in assessment report

[iantaylor-NOAA]

Problem

The individual functions in sa4ss are really well documented, but I think users would benefit from more information on the bigger picture workflow after creating the skeleton via sa4ss::draft() as described in the README.

I've looked over the 2021 Dover, Lingcod, and Vermillion repositories and see many similarities but also various differences. I appreciate that sa4ss is more like a toolbox than a architectural plan, but I think building on each others' assessments would be easier if we could align some of our basic choices.

Proposal

• agree on some trivial organizational details (if possible):
  • figures directory within the doc (a.k.a. write_up) directory or parallel to it?
  • tables for base model in tables or tex_tables and in model directory or in doc?
  • don't clutter your github repo with model files like I did with lingcod (use Google Drive instead unless somebody has a better idea, and are we allowed or is it a good idea to use the {googledrive} package)?
• provide some general guidance on workflow via an expanded README or a vignette:
  • how to set up the file structure agreed above
  • when to run read_model()
  • info on adding figures
    • basic Markdown as in the skeleton,
    • via sa4ss::add_figure(), or
    • via custom wrapper for all the r4ss plots like lingcod::add_figure_ling()
  • what steps to take each time you render the PDF, such as via the following script and function
    • https://github.com/chantelwetzel-noaa/dover_sole_2021/blob/master/write_up/_create_doc.R
    • https://github.com/pfmc-assessments/lingcod/blob/main/R/compile_ling.R

I'm happy to help with any of these things but I don't know enough about the package to do much without some consensus from the more experienced folks.