CalibrationDiagnosticSuite#

class rojak.turbulence.diagnostic.CalibrationDiagnosticSuite(factory, diagnostics)[source]#

A DiagnosticSuite computed on a calibration dataset, used to derive severity thresholds and EDR distribution parameters for an EvaluationDiagnosticSuite

Parameters:
__init__(factory, diagnostics)[source]#

See DiagnosticSuite.__init__()

Parameters:
Return type:

None

Methods

__init__(factory, diagnostics)

See DiagnosticSuite.__init__()

as_dataset()

Combine every diagnostic's computed value into a single Dataset

compute_distribution_parameters()

Compute the log-normal distribution histogram for every diagnostic in the suite

compute_thresholds(percentile_config)

Compute the percentile-based severity thresholds for every diagnostic in the suite

computed_values(progress_description)

Iterate over the name and computed value of each diagnostic in the suite

computed_values_as_dict()

Same as computed_values(), collected eagerly into a dict (without a progress bar)

diagnostic_names()

Names of the diagnostics in the suite

export_as_zarr(output_path, /, *[, zarr_format])

Export every diagnostic's computed value (see as_dataset()) to a zarr store

get_prototype_computed_diagnostic()

An arbitrary diagnostic's computed value, used where any one diagnostic's shape/coords will do

load_from_zarr(path, target_diagnostics, *)

Construct a DiagnosticSuite from a zarr store of previously-exported diagnostic values