dival.measure module

Provides the abstract Measure base class and some popular measures.

Measure instances are identified by a unique short_name, which is used by the evaluation module.

dival.measure.gen_unique_name(name_orig)

class dival.measure.Measure(short_name=None)

Bases: abc.ABC

Abstract base class for measures used for evaluation.

In subclasses, either __init__() should be inherited or it should call super().__init__() in order to register the short_name and to ensure it is unique.

measure_type

The measure type. Measures with type 'distance' should attain small values if the reconstruction is good. Measures with type 'quality' should attain large values if the reconstruction is good.

Type

{'distance', 'quality'}

short_name

Short name of the measure, used as identifier (key in measure_dict).

Type

str

name

Name of the measure.

Type

str

description

Description of the measure.

Type

str

measure_type = None

Class attribute, default value for measure_type.

name = ''

Class attribute, default value for name.

description = ''

Class attribute, default value for description.

measure_dict = {'l2': <dival.measure.L2Measure object>, 'mse': <dival.measure.MSEMeasure object>, 'psnr': <dival.measure.PSNRMeasure object>, 'ssim': <dival.measure.SSIMMeasure object>}

Class attribute, registry of all measures with their short_name as key.

__init__(short_name=None)

Parameters

short_name (str, optional) – The short name of this measure, used as identifier in measure_dict. If None is passed and short_name was not set by the subclass, the __name__ of the subclass is used. If short_name is already taken by another instance, a unique short name is generated by appending a suffix of format '_%d'.

short_name = ''

Class attribute, default value for short_name.

abstract apply(reconstruction, ground_truth)

Calculate the value of this measure.

Parameters

• reconstruction (odl element) – The reconstruction.

• ground_truth (odl element) – The ground truth to compare with.

Returns

value – The value of this measure for the given reconstruction and ground_truth.

Return type

float

classmethod get_by_short_name(short_name)

Return Measure instance with given short name by registry lookup.

Parameters

short_name (str) – Short name, identifier in measure_dict.

Returns

measure – The instance.

Return type

Measure

as_operator_for_fixed_ground_truth(ground_truth)

Return an odl operator that can be applied to different reconstructions for fixed ground truth.

Returns

op – odl operator.

Return type

odl operator

class dival.measure.L2Measure(short_name=None)

Bases: dival.measure.Measure

The euclidean (l2) distance measure.

measure_type = 'distance'

short_name = 'l2'

name = 'euclidean distance'

description = 'distance given by sqrt(sum((reconstruction-ground_truth)**2))'

apply(reconstruction, ground_truth)

Calculate the value of this measure.

Parameters

• reconstruction (odl element) – The reconstruction.

• ground_truth (odl element) – The ground truth to compare with.

Returns

value – The value of this measure for the given reconstruction and ground_truth.

Return type

float

class dival.measure.MSEMeasure(short_name=None)

Bases: dival.measure.Measure

The mean squared error distance measure.

measure_type = 'distance'

short_name = 'mse'

name = 'mean squared error'

description = 'distance given by 1/n * sum((reconstruction-ground_truth)**2)'

apply(reconstruction, ground_truth)

Calculate the value of this measure.

Parameters

• reconstruction (odl element) – The reconstruction.

• ground_truth (odl element) – The ground truth to compare with.

Returns

value – The value of this measure for the given reconstruction and ground_truth.

Return type

float

dival.measure.MSE = <dival.measure.MSEMeasure object>

Instance of MSEMeasure, with short_name='mse'.

class dival.measure.PSNRMeasure(data_range=None, short_name=None)

Bases: dival.measure.Measure

The peak signal-to-noise ratio (PSNR) measure.

The data range is automatically determined from the ground truth if not given to the constructor.

data_range

The data range (max-min possible value). If data_range is None, np.max(ground_truth) - np.min(ground_truth) is used in apply().

Type

float or None

measure_type = 'quality'

short_name = 'psnr'

name = 'peak signal-to-noise ratio'

description = 'quality given by 10*log10(MAX**2/MSE)'

__init__(data_range=None, short_name=None)

Parameters

• data_range (float, optional) – The data range (max-min possible value). If data_range is None, np.max(ground_truth) - np.min(ground_truth) is used in apply().

• short_name (str, optional) – Short name.

apply(reconstruction, ground_truth)

Calculate the value of this measure.

Parameters

• reconstruction (odl element) – The reconstruction.

• ground_truth (odl element) – The ground truth to compare with.

Returns

value – The value of this measure for the given reconstruction and ground_truth.

Return type

float

dival.measure.PSNR = <dival.measure.PSNRMeasure object>

Instance of PSNRMeasure, with data_range=None and short_name='psnr'.

class dival.measure.SSIMMeasure(short_name=None, **kwargs)

Bases: dival.measure.Measure

The structural similarity index measure.

measure_type = 'quality'

short_name = 'ssim'

name = 'structural similarity index'

description = 'The (M)SSIM like described in `Wang et al. 2014 <https://doi.org/10.1109/TIP.2003.819861>`_.'

__init__(short_name=None, **kwargs)

This is a wrapper for skimage.metrics.structural_similarity(). The data range is automatically determined from the ground truth if not given to the constructor.

Parameters

• short_name (str, optional) – Short name.

• kwargs (dict, optional) – Keyword arguments that will be passed to structural_similarity() in apply(). If data_range is not specified, np.max(ground_truth) - np.min(ground_truth) is used.

apply(reconstruction, ground_truth)

Calculate the value of this measure.

Parameters

• reconstruction (odl element) – The reconstruction.

• ground_truth (odl element) – The ground truth to compare with.

Returns

value – The value of this measure for the given reconstruction and ground_truth.

Return type

float

dival.measure.SSIM = <dival.measure.SSIMMeasure object>

Instance of SSIMMeasure, with data_range=None and short_name='ssim'.