report

report ¶

ValidationReport dataclass for DRO comparison results.

This module provides the data container for storing validation comparison results against OSIPI Digital Reference Objects or other reference datasets.

References

QIBA DCE Profile v2.0 OSIPI Task Force 4.1/4.2 publications

ValidationReport `dataclass` ¶

ValidationReport(
    reference_name,
    reference_parameters,
    computed_parameters,
    absolute_errors=dict(),
    relative_errors=dict(),
    tolerances=dict(),
    within_tolerance=dict(),
    pass_rate=dict(),
    overall_pass=False,
)

Validation comparison against reference values.

This dataclass contains the results of comparing computed parameter maps against reference (ground truth) values from OSIPI DROs or other validation datasets.

ATTRIBUTE	DESCRIPTION
`reference_name`	Name of reference dataset (e.g., 'OSIPI_DRO_v10'). TYPE: `str`
`reference_parameters`	Ground truth parameter values. TYPE: `dict[str, NDArray[floating]]`
`computed_parameters`	osipy computed parameter values. TYPE: `dict[str, NDArray[floating]]`
`absolute_errors`	Absolute errors \|computed - reference\|. TYPE: `dict[str, NDArray[floating]]`
`relative_errors`	Relative errors \|computed - reference\| / reference. TYPE: `dict[str, NDArray[floating]]`
`tolerances`	Tolerance thresholds for each parameter. Format: {'Ktrans': {'absolute': 0.005, 'relative': 0.01}} TYPE: `dict[str, dict[str, float]]`
`within_tolerance`	Boolean mask indicating voxels within tolerance. TYPE: `dict[str, NDArray[bool_]]`
`pass_rate`	Fraction of voxels within tolerance per parameter. TYPE: `dict[str, float]`
`overall_pass`	True if all parameters meet their tolerance criteria. TYPE: `bool`

Examples:

>>> import numpy as np
>>> from osipy.common.validation.report import ValidationReport
>>> ref = {"Ktrans": np.array([0.1, 0.2, 0.3])}
>>> comp = {"Ktrans": np.array([0.105, 0.195, 0.31])}
>>> report = ValidationReport(
...     reference_name="OSIPI_DRO_v10",
...     reference_parameters=ref,
...     computed_parameters=comp,
...     absolute_errors={"Ktrans": np.abs(comp["Ktrans"] - ref["Ktrans"])},
...     relative_errors={"Ktrans": np.abs(comp["Ktrans"] - ref["Ktrans"]) / ref["Ktrans"]},
...     tolerances={"Ktrans": {"absolute": 0.01, "relative": 0.05}},
...     within_tolerance={"Ktrans": np.array([True, True, True])},
...     pass_rate={"Ktrans": 1.0},
...     overall_pass=True,
... )

parameters `property` ¶

parameters

Return list of validated parameter names.

n_parameters `property` ¶

n_parameters

Return number of validated parameters.

summary ¶

summary()

Generate human-readable summary of validation results.

RETURNS	DESCRIPTION
`str`	Formatted summary string.

get_statistics ¶

get_statistics(parameter)

Get error statistics for a specific parameter.

PARAMETER	DESCRIPTION
`parameter`	Parameter name. TYPE: `str`

RETURNS	DESCRIPTION
`dict[str, float]`	Statistics including mean/std/max absolute and relative errors.

to_dict ¶

to_dict()

Serialize report to a dictionary.

RETURNS	DESCRIPTION
`dict[str, Any]`	Dict with keys: reference_name, overall_pass, parameters (dict of param -> {pass_rate, statistics}), tolerances, timestamp, version.

to_json ¶

to_json(path=None)

Serialize report to JSON string.

PARAMETER	DESCRIPTION
`path`	Optional file path. If given, writes JSON to file. TYPE: `str \| Path \| None` DEFAULT: `None`

RETURNS	DESCRIPTION
`str`	JSON string.