Deborah.Rebekah.SummaryLoader

load_summary(
    work::String,
    analysis_ensemble::String,
    cumulant_name::String,
    overall_name::String,
    keys::Vector{String},
    labels::Vector{String},
    trains::Vector{String}
) -> Dict{String, Array{Float64,2}}

Load summary statistics (mean and error) from .dat files produced by Deborah.Esther or Deborah.DeborahCore.

Each summary file is expected to contain comment lines in the format: #<key> <mean> <stderr>

These lines are parsed into 2D matrices indexed by (label, train) combinations, and stored as:

• "key:avg" → matrix of means
• "key:err" → matrix of errors

Arguments

• work::String : Root working directory
• analysis_ensemble::String : Ensemble directory (e.g., "L8T4b1.60k13570")
• cumulant_name::String : Observable subfolder (e.g., "cond", "susp")
• overall_name::String : Base name used in filenames
• keys::Vector{String} : Observable key(s), usually one or more of "cond", "susp", or "Deborah"
• labels::Vector{String} : Labeled set ratio tags (LBP, row index)
• trains::Vector{String} : Training set ratio tags (TRP, column index)

Returns

• Dict{String, Array{Float64,2}} : Dictionary with keys like "cond:avg", "cond:err" or "Deborah:avg", each mapped to a matrix of shape (length(labels), length(trains))