Users currently construct rfaR inputs by copy-pasting from RMC-BestFit into CSV files, which is error-prone — copied data can be silently mangled, numeric columns can lose type, date formats can shift.
Reading directly from RMC-BestFit's .sqlite output would eliminate the copy-paste step entirely: no manual transcription, no Excel mangling, deterministic field types.
Scope
Long-term enhancement. Design work needed:
- Identify the relevant tables/views in RMC-BestFit's SQLite schema
- Design a reader function (e.g.,
read_bestfit_params(), read_bestfit_hydrographs()) that returns rfaR-ready data frames
- Decide how to handle versioning if the SQLite schema changes across BestFit releases
Feedback source
Raised by Allen (comment #9). Would also largely address comment #6 (CSV copy-paste corruption).
Users currently construct rfaR inputs by copy-pasting from RMC-BestFit into CSV files, which is error-prone — copied data can be silently mangled, numeric columns can lose type, date formats can shift.
Reading directly from RMC-BestFit's
.sqliteoutput would eliminate the copy-paste step entirely: no manual transcription, no Excel mangling, deterministic field types.Scope
Long-term enhancement. Design work needed:
read_bestfit_params(),read_bestfit_hydrographs()) that returns rfaR-ready data framesFeedback source
Raised by Allen (comment #9). Would also largely address comment #6 (CSV copy-paste corruption).