Deborah.DeborahDocument

module DeborahDocument

Deborah.DeborahDocument — Postprocessing and document generation for Deborah.DeborahCore results.

The Deborah.DeborahDocument module provides a reporting pipeline on top of Deborah.DeborahCore. It loads summary results, archives the results into JLD2 format. This module is designed as the "document layer" in the Deborah.jl ecosystem, complementing Deborah.DeborahCore (ML + bias correction) with a reporting backend.

Scope & Responsibilities

• Configuration: Parse a TOML config for data, bootstrap/jackknife, and abbreviation maps.
• Summary loading: Collect pre-computed results via Deborah.Rebekah.SummaryLoader.
• Archiving: Save results into .jld2 format with Deborah.Rebekah.JLD2Saver, and copy snapshots to the working directory.
• Integration: Connects Deborah.Sarah utilities (logging, name parsing, transcoding) and Deborah.Rebekah's reporting tools.

Key Component

• DeborahDocumentRunner.run_DeborahDocument Main entry point. Runs the document-generation workflow for a single ensemble.

Typical Outputs

• Results snapshot: results_<overall_name>.jld2 in analysis directory.

Minimal Usage

julia> using Deborah
julia> run_DeborahDocument("config_DeborahThreads.toml")

Notes

• Abbreviation dictionary [abbreviation] in the TOML is parsed with Deborah.Sarah.StringTranscoder.parse_string_dict.
• Input/output codes (XY_code, X_Y) and model suffixes are built with Deborah.Sarah.NameParser.
• Assumes upstream Deborah.DeborahCore runs have produced summary data available for loading.

See Also

• Deborah.DeborahCore — core ML/bias-correction pipeline.
• Deborah.Rebekah — reporting/plotting backend.