Link Search Menu Expand Document

Curves

Table of contents

analyze_curve()

It provides an overall analysis of the model performances by plotting the desired curve.

Parameters

curve

Curves, optional
Evaluation curve used for the analysis.
(default is Curves.PRECISION_RECALL_CURVE)
average

str, optional
Indicates the averaging method. It can be 'macro' or 'micro'.
(default is 'macro')
show

bool, optional
Indicates whether the plot should be shown or not. If False, returns the results as dict.
(default is True)

Example

Classification

from odin.classes import AnalyzerClassification

my_analyzer = AnalyzerClassification("my_classifier_name", my_classification_dataset)
my_analyzer.analyze_curve()

Localization

from odin.classes import AnalyzerLocalization

my_analyzer = AnalyzerLocalization("my_detector_name", my_localization_dataset)
my_analyzer.analyze_curve()

analyze_curve_output


Models comparison

Example

Classification

from odin.classes import ComparatorClassification

my_comparator = ComparatorClassification(dataset_gt_param, classification_type, models_proposals)
my_comparator.analyze_curve()

Localization

from odin.classes import ComparatorLocalization

my_comparator = ComparatorLocalization(dataset_gt_param, task_type, models_proposals)
my_comparator.analyze_curve()

compare_models_on_curve_output


Tasks supported

Binary Classification Single-label Classification Multi-label Classification Object Detection Instance Segmentation
yes yes yes yes yes

analyze_curve_for_categories()

For each category, it provides an analysis of the model performances by plotting the desired curve.

Parameters

categories

list, optional
List of categories to be included in the analysis. If not specified, all the categories are included.
(default is None)
curve

Curves, optional
Evaluation curve used for the analysis.
(default is Curves.PRECISION_RECALL_CURVE)
show

bool, optional
Indicates whether the plot should be shown or not. If False, returns the results as dict.
(default is True)

Example

Classification

from odin.classes import AnalyzerClassification

my_analyzer = AnalyzerClassification("my_classifier_name", my_classification_dataset)
my_analyzer.analyze_curve_for_categories()

Localization

from odin.classes import AnalyzerLocalization

my_analyzer = AnalyzerLocalization("my_detector_name", my_localization_dataset)
my_analyzer.analyze_curve_for_categories()

analyze_curve_for_categories_output


Models comparison

Example

Classification

from odin.classes import ComparatorClassification

my_comparator = ComparatorClassification(dataset_gt_param, classification_type, models_proposals)
my_comparator.analyze_curve_for_categories()

Localization

from odin.classes import ComparatorLocalization

my_comparator = ComparatorLocalization(dataset_gt_param, task_type, models_proposals)
my_comparator.analyze_curve_for_categories()

compare_models_on_curve_output


Tasks supported

Binary Classification Single-label Classification Multi-label Classification Object Detection Instance Segmentation
no yes yes yes yes