Deborah.Sarah.SummaryCollector

Deborah.Sarah.SummaryCollector.collect_summaries_Deborah — Method

collect_summaries_Deborah(
    trace_data::Dict{String, Vector{Vector{T}}},
    bootstrap_data::Dict{Symbol, Any},
    bin_size::Int
) -> Tuple{Dict{String, Tuple{T, T}}, Dict{String, Tuple{T, T}}} where T<:Real

Collect jackknife and bootstrap summary statistics for Deborah.DeborahCore model output.

Arguments

trace_data::Dict{String, Vector{Vector{T}}}: Raw trace data, where "Y_info" key holds a vector of vectors.
bootstrap_data::Dict{Symbol, Any}: Contains precomputed bootstrap means under :mean field.
bin_size::Int: Bin size to use for jackknife resampling.

Returns

(summary_jackknife, summary_bootstrap):
- Dict{String, Tuple{T, T}}: Maps observable names to (mean, error) from jackknife.
- Dict{String, Tuple{T, T}}: Maps observable names to (mean, error) from bootstrap.

Notes

Jackknife is computed only for "Y_info" (first vector in the list).
Bootstrap summaries are computed for all observables in the suffix list.
Uses Jackknife.jackknife_average_error_from_raw and Bootstrap.bootstrap_average_error internally.

source

Deborah.Sarah.SummaryCollector.collect_summaries_Esther — Method

collect_summaries_Esther(
    trace_data::Dict{String, Vector{Vector{T}}},
    Q_moment::Dict{String, Vector{T}},
    jackknife_data::Dict{Symbol, Any},
    bootstrap_data::Dict{Symbol, Any},
    bin_size::Int
) -> Tuple{Dict{String, Tuple{T, T}}, Dict{String, Tuple{T, T}}} where T<:Real

Collect jackknife and bootstrap summary results for the Deborah.Esther model into two dictionaries.

Arguments

trace_data::Dict{String, Vector{Vector{T}}}: Raw trace data grouped by trM1-trM4.
Q_moment::Dict{String, Vector{T}}: Moment-transformed trace data grouped by Q1-Q4.
jackknife_data::Dict{Symbol, Any}: Precomputed jackknife means under :mean.
bootstrap_data::Dict{Symbol, Any}: Precomputed bootstrap means under :mean.
bin_size::Int: Bin size for jackknife error estimation.

Returns

(summary_jackknife, summary_bootstrap):
- Dict{String, Tuple{T, T}}: Maps each observable (e.g., "trM1:Y_info") to a tuple of (mean, error) for jackknife.
- Dict{String, Tuple{T, T}}: Same structure for bootstrap results.

Notes

Uses Jackknife.jackknife_average_error_from_raw, Jackknife.jackknife_average_error, and Bootstrap.bootstrap_average_error internally to extract (mean, error) tuples.
Jackknife is only computed for Y_info; bootstrap is computed for all suffixes.

source