def test_filter_container_annotations_and_container_predictions_by_labels_with_remove_using_apply_to(self):
config = [{'type': 'filter', 'apply_to': 'all', 'labels': ['to_be_filtered'], 'remove_filtered': True}]
prediction = ContainerPrediction({'prediction': DetectionPrediction(labels=['some_label', 'to_be_filtered'])})
expected_prediction = ContainerPrediction({'prediction': DetectionPrediction(labels=['some_label'])})
annotation = ContainerAnnotation({'annotation': DetectionAnnotation(labels=['some_label', 'to_be_filtered'])})
expected_annotation = ContainerAnnotation({'annotation': DetectionAnnotation(labels=['some_label'])})
postprocess_data(PostprocessingExecutor(config), [annotation], [prediction])
assert prediction == expected_prediction and annotation == expected_annotation
def test_filter_predictions_by_labels_with_remove_on_container_using_apply_to(self):
config = [{'type': 'filter', 'apply_to': 'prediction', 'labels': ['to_be_filtered'], 'remove_filtered': True}]
prediction = ContainerPrediction(
{'detection_out': DetectionPrediction(labels=['some_label', 'to_be_filtered'])}
)
expected = ContainerPrediction({'detection_out': DetectionPrediction(labels=['some_label'])})
postprocess_data(PostprocessingExecutor(config), [None], [prediction])
assert prediction == expected
def test_filter_regular_annotations_and_container_predictions_by_labels_with_ignore_using_apply_to(self):
config = [{'type': 'filter', 'apply_to': 'all', 'labels': ['to_be_filtered'], 'remove_filtered': False}]
prediction = ContainerPrediction(
{'detection_out': DetectionPrediction(labels=['some_label', 'to_be_filtered'])})
expected_prediction = ContainerPrediction(
{'detection_out': DetectionPrediction(labels=['some_label', 'to_be_filtered'],
metadata={'difficult_boxes': [1]})})
annotation = DetectionAnnotation(labels=['some_label', 'to_be_filtered'])
expected_annotation = DetectionAnnotation(labels=['some_label', 'to_be_filtered'],
metadata={'difficult_boxes': [1]})
postprocess_data(PostprocessingExecutor(config), [annotation], [prediction])
assert prediction == expected_prediction and annotation == expected_annotation
def test_filter_predictions_by_labels_with_ignore_on_container(self):
config = [{'type': 'filter', 'prediction_source': 'detection_out',
'labels': ['to_be_filtered'], 'remove_filtered': False}]
prediction = ContainerPrediction(
{'detection_out': DetectionPrediction(labels=['some_label', 'to_be_filtered'])}
)
expected = ContainerPrediction(
{'detection_out': DetectionPrediction(labels=['some_label', 'to_be_filtered'],
metadata={'difficult_boxes': [1]})}
)
postprocess_data(PostprocessingExecutor(config), [None], [prediction])
assert prediction == expected
def test_filter_multi_source_predictions_by_labels_with_remove(self):
config = [{'type': 'filter', 'prediction_source': ['detection_out1', 'detection_out2'],
'labels': ['to_be_filtered'], 'remove_filtered': True}]
prediction = ContainerPrediction(
{'detection_out1': DetectionPrediction(labels=['some_label', 'to_be_filtered']),
'detection_out2': DetectionPrediction(labels=['some_label', 'to_be_filtered'])}
)
expected = ContainerPrediction(
{'detection_out1': DetectionPrediction(labels=['some_label']),
'detection_out2': DetectionPrediction(labels=['some_label'])}
)
postprocess_data(PostprocessingExecutor(config), [None], [prediction])
assert prediction == expected
def test_complete_accuracy_with_container_sources(self):
annotations = [
ContainerAnnotation(
{'a': ClassificationAnnotation('identifier', 3)})
]
predictions = [
ContainerPrediction({
'p':
ClassificationPrediction('identifier', [1.0, 1.0, 1.0, 4.0])
})
]
config = [{
'type': 'accuracy',
'top_k': 1,
'annotation_source': 'a',
'prediction_source': 'p'
}]
dispatcher = MetricsExecutor(config, None)
dispatcher.update_metrics_on_batch(range(len(annotations)),
annotations, predictions)
for _, evaluation_result in dispatcher.iterate_metrics(
annotations, predictions):
assert evaluation_result.name == 'accuracy'
assert evaluation_result.evaluated_value == pytest.approx(1.0)
assert evaluation_result.reference_value is None
assert evaluation_result.threshold is None
def test_accuracy_with_unsupported_prediction_type_as_prediction_source_for_container_raises_config_error(self):
annotations = [ClassificationAnnotation('identifier', 3)]
predictions = [ContainerPrediction({'prediction': DetectionPrediction('identifier', [1.0, 1.0, 1.0, 4.0])})]
dispatcher = MetricsExecutor([{'type': 'accuracy', 'top_k': 1, 'prediction_source': 'prediction'}], None)
with pytest.raises(ConfigError):
dispatcher.update_metrics_on_batch(annotations, predictions)
def test_filter_predictions_unsupported_source_type_raise_type_error_exception(self):
config = [{'type': 'filter', 'prediction_source': 'detection_out', 'labels': ['to_be_filtered'],
'remove_filtered': False}]
prediction = ContainerPrediction({'detection_out': ClassificationAnnotation()})
executor = PostprocessingExecutor(config)
with pytest.raises(TypeError):
postprocess_data(executor, [None], [prediction])
def test_accuracy_with_unsupported_prediction_in_container_raise_config_error_exception(self):
annotations = [ClassificationAnnotation('identifier', 3)]
predictions = [ContainerPrediction({'prediction': DetectionPrediction('identifier', [1.0, 1.0, 1.0, 4.0])})]
config = {'annotation': 'mocked', 'metrics': [{'type': 'accuracy', 'top_k': 1}]}
dispatcher = MetricsExecutor(config, None)
with pytest.raises(ConfigError):
dispatcher.update_metrics_on_batch(annotations, predictions)
def test_filter_predictions_source_not_found_raise_config_error_exception(self):
config = [{'type': 'filter', 'prediction_source': 'undefined', 'labels': ['to_be_filtered']}]
prediction = ContainerPrediction(
{'detection_out': DetectionPrediction(labels=['some_label', 'to_be_filtered'])}
)
executor = PostprocessingExecutor(config)
with pytest.raises(ConfigError):
postprocess_data(executor, [None], [prediction])
def test_accuracy_on_prediction_container_with_several_suitable_representations_raise_config_error_exception(self):
annotations = [ClassificationAnnotation('identifier', 3)]
predictions = [ContainerPrediction({
'prediction1': ClassificationPrediction('identifier', [1.0, 1.0, 1.0, 4.0]),
'prediction2': ClassificationPrediction('identifier', [1.0, 1.0, 1.0, 4.0])
})]
dispatcher = MetricsExecutor([{'type': 'accuracy', 'top_k': 1}], None)
with pytest.raises(ConfigError):
dispatcher.update_metrics_on_batch(annotations, predictions)
