Reorganize downscaling workflow & generate design points for herbaceous crops

[dlebauer]

This PR does continues to update and reorganize the downscaling workflow while also generating design points for herbaceous crops.

Apologies that it didn't seem practical to separate these steps into separate PRs at the time.

Key Changes

• Improvements to reproducibility, logging, and workflow clarity
• Update documentation, comments, logging throughout the codebase for clarity
• Renamed and renumbered workflow files to make it easier to split workflows into separate steps (e.g. using 010, 011, 020) nothing is future proof; can be revisited in the future.
• Centralize configuration in 000-config.R
  • input/ output directories and file paths
  • relative to CCMMF_DIR set in .Renviron
  • configure whether it is a PRODUCTION run, if FALSE, it is faster
    • added associated conditionals for slow steps
• Combine woody and herbaceous design points into single design_points.csv, added 'pft' field
• More consistent variable names
• Add LandIQ data summaries to 009_update_landiq.R
• Update county-level aggregation and plotting in 040_downscale_and_aggregate.R
• Add ALE and ICE plot generation in 041_downscale_analysis.R; add plots to reports/downscaling_results.qmd
• Fix minor bugs and improve code style (e.g., consistent use of pkg::function() syntax)
• Rename data_raw/anchor_sites.csv --> data_raw/anchor_site_locations.csv and data/anchor_sites_ids.csv --> anchor_sites.csv as recommended in 1b downscaling workflow workflows#2

Review Request

Please review for correctness. Feedback on this reorganization is welcome, where possible I would prefer to address these in subsequent PRs rather than holding this one up.

Compiled report for review: downscaling_results.zip

[infotroph]

prepare.qmd no longer exists

[dlebauer]

I made some requested changes, defered most to next iteration - enumerated in ccmmf/organization#145

[dlebauer]

What's been improved in the calculation since 1b? How have the county maps changed or not?
My first thought is the what's-new would go in workflow_documentation.md and the discussion of what's changed would go in downscaling_results.qmd.

Great points. I will keep workflow documentation (how) and results (what) focused on the current state and then add a CHANGELOG.md + release-notes so we can track what’s new across phases. Phase 2b will compare herbaceous vs woody.

Also, I reconciled design_points.csv in 1cb2f61

[infotroph]

Not that you necessarily want a third option, but in the workflows repo I've converged on passing all configuration variables as command-line arguments with defaults set via {optparse}. It's verbose but highly discoverable, allows overrides for any one step if needed, and forced me to actually write the help strings for everything.

[infotroph]

I see one missing comment char that'll be an error until fixed, otherwise 👍