Deborah.DeborahEstherMiriam

module DeborahEstherMiriam

Deborah.DeborahEstherMiriam — Full pipeline integration: Deborah.DeborahCore → Deborah.Esther → Deborah.Miriam.

The Deborah.DeborahEstherMiriam module provides the highest-level orchestration layer in the Deborah.jl ecosystem. It unifies Deborah.DeborahCore (ML-based trace estimation), Deborah.Esther (cumulant analysis at the single ensemble), and Deborah.Miriam (multi-ensemble reweighting and interpolation). This module guarantees that all intermediate results exist (creating them on demand), and drives single-ensemble-level and multi-ensemble analyses end-to-end.

Scope & Responsibilities

• Single-ensemble prerequisites:
  • MiriamDependencyManager.ensure_ensemble_exists Ensures that each ensemble has complete Deborah.DeborahCore $+$ Deborah.Esther outputs, invoking MiriamDependencyManager.run_Esther_from_Miriam where needed.
• Multi-ensemble prerequisites:
  • MiriamExistenceManager.ensure_multi_ensemble_exists Checks whether required .dat outputs for combined ensembles exist; if missing, re-runs ensemble-level Deborah.Esther $+$ Deborah.Miriam steps automatically.
• Full pipeline execution:
  • DeborahEstherMiriamRunner.run_Deborah_Esther_Miriam Entry point for running the complete Deborah.DeborahCore → Deborah.Esther → Deborah.Miriam process.

Workflow

1. User provides a multi-ensemble TOML config with data paths, bootstrap/jackknife parameters, and model specifications.
2. MiriamDependencyManager.ensure_ensemble_exists validates or generates Deborah.DeborahCore/Deborah.Esther outputs for each ensemble.
3. MiriamExistenceManager.ensure_multi_ensemble_exists validates or regenerates multi-ensemble Miriam outputs.
4. DeborahEstherMiriamRunner.run_Deborah_Esther_Miriam executes the Deborah.Miriam stage, producing multi-ensemble-reweighted and interpolated cumulants across ensembles.

Outputs

• Single-ensemble-level results: summary_Esther_*.dat, trace-based bootstrap/jackknife bundles.
• Multi-ensemble results: RWBS, RWJK, RWP1, RWP2, Y_BS, Y_JK, Y_P1, Y_P2 .dat files in analysis directories.
• Logs: Structured job-aware logging (Deborah.Sarah.JobLoggerTools) during each stage.

Minimal Usage

julia> using Deborah
julia> run_Deborah_Esther_Miriam("config_Miriam.toml")

Notes

• Abbreviation maps are supported for compact path encoding.
• Both single-ensemble and multi-ensemble runs are guarded: missing files trigger regeneration automatically.
• Bootstrap (N_bs, blk_size, method) and jackknife (bin_size) configs are consistently propagated across Deborah, Esther, and Miriam layers.
• All path/filename construction respects LBP/TRP splits and ensemble tags.

See Also

• Deborah.DeborahCore — ML trace estimation.
• Deborah.DeborahEsther — bridge from Deborah.DeborahCore to Deborah.Esther.
• Deborah.Miriam — multi-ensemble reweighting & interpolation.