Tidy the per-fit diagnostics from a calibration result

Extracts sampler/optimizer diagnostics into a tidy frame keyed by curve_id. Bayesian columns (rhat_max, ess_bulk_min, ess_tail_min, n_divergent, pct_divergent, max_treedepth_hit, ebfmi_min) and frequentist columns (hessian_condition_number, gradient_norm, optimizer_code, rel_tol_achieved) coexist; the irrelevant set is NA for a given engine. Common columns: fit_seconds, n_iterations, converged, fit_seed. Feeds calib_fit_diag.

Usage

tidy_fit_diag(x, ...)

# S3 method for class 'calibration_result'
tidy_fit_diag(x, ...)

# S3 method for class 'calibration_result_multiplate'
tidy_fit_diag(x, ...)

Arguments

x

A calibration_result or calibration_result_multiplate.

...

Unused; for method extensibility.

Value

One row per curve_id.

See also

tidy_hyperparam()