Create a Mermaid diagram in docs/architecture/solver_data_pipeline.md showing the end-to-end flow: config (YAML or DB) → data sources (potential locations, census demographics, driving/haversine distances) → model_data.py assembly → KP solver (model_run.py) → results output (CSV or DB). Show where census_data_type branches (redistricting/CVAP/etc.) and where driving vs. haversine distance is chosen.
Before drawing: work with Claude Code to identify the specific functions and files to anchor each node on, and to settle the right level of abstraction (how many boxes, what's in/out of scope) — don't rely on this issue description alone, the codebase may have moved on by the time this is picked up.
Create a Mermaid diagram in
docs/architecture/solver_data_pipeline.mdshowing the end-to-end flow: config (YAML or DB) → data sources (potential locations, census demographics, driving/haversine distances) →model_data.pyassembly → KP solver (model_run.py) → results output (CSV or DB). Show wherecensus_data_typebranches (redistricting/CVAP/etc.) and where driving vs. haversine distance is chosen.Before drawing: work with Claude Code to identify the specific functions and files to anchor each node on, and to settle the right level of abstraction (how many boxes, what's in/out of scope) — don't rely on this issue description alone, the codebase may have moved on by the time this is picked up.