kinextract.estimate_losvd_errors

kinextract.estimate_losvd_errors(fit: dict, cfg, *, run_laplace: bool = True, run_bootstrap: bool = True, run_bias: bool = True, n_bootstrap: int = 200, block_size: int = 1, n_jobs: int = 1, bootstrap_seed: int = 42, confidence: float = 0.68, plot: bool = True, write_to_files: bool = False, prefix: str | None = None, outdir: str | Path = '.') → dict[source]

Convenience wrapper: run all requested error estimation methods and return the combined summary.

Parameters:

  • fit (output of run_spectral_fit())

  • cfg (FitConfig used for the original fit)

  • run_laplace (compute Laplace posterior covariance)

  • run_bootstrap (run residual bootstrap)

  • run_bias (compute bias-corrected LOSVD)

  • n_bootstrap (number of bootstrap replicates)

  • block_size (block size for correlated-noise bootstrap)

  • n_jobs (parallel workers for bootstrap (-1 = all CPUs))

  • bootstrap_seed (RNG seed)

  • confidence (confidence level for intervals (0.68 ≈ 1-sigma))

  • plot (show plots when done)

  • write_to_files (write error estimates to .losvd_errs.out, .gh_errs.out files)

  • prefix (output file prefix (required if write_to_files=True))

  • outdir (output directory for files (default: current directory))

Returns:

summary –

If write_to_files=True, also includes keys:

’losvd_errs_file’: path to {prefix}.losvd_errs.out ‘gh_errs_file’: path to {prefix}.gh_errs.out ‘bootstrap_diag_file’: path to {prefix}.bootstrap_diag.txt (if bootstrap run)

Return type:

dict (see LOSVDErrorEstimator.summarize)