Deborah.Miriam.WriteBSOutput

Deborah.Miriam.WriteBSOutput.write_bs_cumulants — Method

write_bs_cumulants(
    ens_bundle::Ensemble.EnsembleArrayBundle{T},
    N_bs::Int,
    blk_size::Int,
    method::String,
    fname_OG::String,
    fname_P1::String,
    fname_P2::String,
    jobid::Union{Nothing, String}=nothing;
    rng_pool::SeedManager.RNGPool=rng_pool,
) where T -> Nothing

Compute and write bootstrap estimates of cumulants for single ensemble.

This function computes three types of cumulants from a given ensemble:

OG : Original data
P1 : Bias-corrected estimator
P2 : weighted average of P1 and labeled set

The output includes bootstrap averages and standard errors for each cumulant: chiral condensate, susceptibility, skewness, kurtosis, and Binder cumulant.

Arguments

ens_bundle::Ensemble.EnsembleArrayBundle{T}: The ensemble data container.
N_bs::Int: Number of bootstrap resamples.
blk_size::Int: Default block size for bootstrap resampling.
method::String : Block-bootstrap scheme (case-sensitive):
- "nonoverlapping" — Nonoverlapping Block Bootstrap (NBB; resample disjoint blocks).
- "moving" — Moving Block Bootstrap (MBB; resample sliding windows).
- "circular" — Circular Block Bootstrap (CBB; sliding windows with wrap-around).
fname_OG::String: Output path for original data.
fname_P1::String: Output path for P1 estimator.
fname_P2::String: Output path for P2 estimator.
jobid::Union{Nothing, String}: Optional job ID for contextual logging.
rng_pool::SeedManager.RNGPool: Optional RNG pool for bootstrap sampling.

Output

Three text files written with bootstrap results: Each line corresponds to a $\kappa$ value with 5 cumulants and their errors.

Notes

The same bootstrap indices are reused across all estimators for consistent comparison.
All output files share the same formatting and cumulant ordering.

source

Deborah.Miriam.WriteBSOutput.write_bs_moments — Method

write_bs_moments(
    ens_bundle::Ensemble.EnsembleArrayBundle{T},
    N_bs::Int,
    blk_size::Int,
    method::String,
    fname_OG::String,
    fname_P1::String,
    fname_P2::String,
    jobid::Union{Nothing, String}=nothing;
    rng_pool::SeedManager.RNGPool=rng_pool,
) where T -> Nothing

Compute and write bootstrap estimates for raw moments ($Q_1$, $Q_2$, $Q_3$, $Q_4$) in the single ensemble.

This function computes three moment estimators from a given ensemble:

OG : Original data
P1 : Bias-corrected estimator
P2 : weighted average of P1 and labeled set

Each estimator is produced via block bootstrap and summarized by its bootstrap mean and standard error for $Q_1$, $Q_2$, $Q_3$, $Q_4$.

Arguments

ens_bundle::Ensemble.EnsembleArrayBundle{T}: Ensemble data container.
N_bs::Int: Number of bootstrap resamples.
blk_size::Int: Default block size for bootstrap resampling.
method::String: Block-bootstrap scheme (case-sensitive), e.g.:
- "nonoverlapping" — Nonoverlapping Block Bootstrap (NBB).
- "moving" — Moving Block Bootstrap (MBB).
- "circular" — Circular Block Bootstrap (CBB).
fname_OG::String: Output path for original data.
fname_P1::String: Output path for P1 estimator.
fname_P2::String: Output path for P2 estimator.
jobid::Union{Nothing, String}: Optional job ID for logging.
rng_pool::SeedManager.RNGPool: RNG pool used for bootstrap.

Output

Three text files written with bootstrap results. Each line contains the $\kappa$ value followed by $Q_1 \pm \sigma_{Q_1}$, $Q_2 \pm \sigma_{Q_2}$, $Q_3 \pm \sigma_{Q_3}$, $Q_4 \pm \sigma_{Q_4}$.

Notes

Internally calls Deborah.Miriam.CumulantsBundle.compute_moments_bundle_raw to generate bootstrap samples for OG/P1/P2 moments.
The same bootstrap index plans are reused across all estimators for consistency.

source

Deborah.Miriam.WriteBSOutput.write_bs_traces — Method

write_bs_traces(
    ens_bundle::Ensemble.EnsembleArrayBundle{T},
    N_bs::Int,
    blk_size::Int,
    method::String,
    fname_OG::String,
    fname_P1::String,
    fname_P2::String,
    jobid::Union{Nothing, String}=nothing;
    rng_pool::SeedManager.RNGPool=rng_pool,
) where T -> Nothing

Compute and write bootstrap estimates for raw trace components ($\text{Tr} \, M^{-n} \; (n=1,2,3,4)$) in the single ensemble.

This function computes three trace estimators from a given ensemble:

OG : Original data
P1 : Bias-corrected estimator
P2 : weighted average of P1 and labeled set

Each estimator is produced via block bootstrap and summarized by its bootstrap mean and standard error for $\text{Tr} \, M^{-n} \; (n=1,2,3,4)$.

Arguments

ens_bundle::Ensemble.EnsembleArrayBundle{T}: Ensemble data container.
N_bs::Int: Number of bootstrap resamples.
blk_size::Int: Default block size for bootstrap resampling.
method::String: Block-bootstrap scheme (case-sensitive), e.g.:
- "nonoverlapping" — Nonoverlapping Block Bootstrap (NBB).
- "moving" — Moving Block Bootstrap (MBB).
- "circular" — Circular Block Bootstrap (CBB).
fname_OG::String: Output path for original data.
fname_P1::String: Output path for P1 estimator.
fname_P2::String: Output path for P2 estimator.
jobid::Union{Nothing, String}: Optional job ID for logging.
rng_pool::SeedManager.RNGPool: RNG pool used for bootstrap.

Output

Three text files written with bootstrap results.

Notes

Traces are taken from un-rescaled vectors via Deborah.Miriam.EnsembleUtils.trMi_rawT before resampling.
Internally calls Deborah.Miriam.CumulantsBundle.compute_traces_bundle_raw to generate bootstrap samples for OG/P1/P2 traces.
The same bootstrap index plans are reused across all estimators for consistency.

source