def test_indirect_matrices_shape_with_missing_sources(self):
dataset_name = "SAMPLE_WARC"
dataset_builder = DatasetBuilder(dataset_name)
dataset_builder.build()
top = dataset_builder.artifacts[0]
middle = dataset_builder.artifacts[1]
bottom = dataset_builder.artifacts[2]
expected_top_middle_shape = (len(top), len(middle))
expected_middle_bottom_shape = (len(middle), len(bottom))
expected_top_bottom_shape = (len(top), len(bottom))
assert_shapes_equal(
dataset_builder.trace_matrix_map["0-1"].matrix, expected_top_middle_shape
)
assert_shapes_equal(
dataset_builder.trace_matrix_map["0-2"].matrix, expected_top_bottom_shape
)
assert_shapes_equal(
dataset_builder.trace_matrix_map["1-2"].matrix,
expected_middle_bottom_shape,
)
self.assertEqual(
dataset_builder.trace_matrix_map["1-2"].matrix.shape,
expected_middle_bottom_shape,
)
def test_creating_missing_matrices(self):
"""
For MockDataset, removes trace matrix 0-2 from structure definition and checks that builder is able to
construct one using transitive traces from the other matrices.
:return:
"""
dataset_builder = DatasetBuilder("MockDataset")
dataset_builder.structure_definition["traces"]["0-2"] = None
dataset_builder.build()
trace_matrix_builder = dataset_builder.trace_matrix_builder
graph = trace_matrix_builder.create_trace_matrix_dependency_graph()
self.assertEqual(
3, len(trace_matrix_builder.trace_matrix_map.get_trace_ids()))
self.assertIsNotNone(trace_matrix_builder.trace_matrix_map["0-2"])
self.assertIsNotNone(
trace_matrix_builder.trace_matrix_map["0-2"].matrix)
self.assertEqual(
1, trace_matrix_builder.trace_matrix_map["0-2"].matrix[0, 0])
self.assertEqual(
2, trace_matrix_builder.trace_matrix_map["0-2"].matrix.sum(
axis=1).sum())
self.assertEqual(3, len(graph.es))
def test_create_transitive_trace_matrices_with_warc(self):
dataset_builder = DatasetBuilder("SAMPLE_WARC")
dataset_builder.build()
trace_matrix_creator = dataset_builder.trace_matrix_builder
graph = (dataset_builder.trace_matrix_builder.
create_trace_matrix_dependency_graph())
self.assertIsNotNone(trace_matrix_creator.trace_matrix_map["1-2"])
self.assertEqual(3, len(graph.es))
def test_get_paths_to_missing_traces_with_fakedataset(self):
dataset_builder = DatasetBuilder("MockDataset")
dataset_builder.structure_definition["traces"]["0-2"] = None
dataset_builder.artifact_builder.build()
dataset_builder.trace_matrix_builder.set_artifact_levels(
dataset_builder.artifact_builder.artifacts)
dataset_builder.trace_matrix_builder.build_matrices_defined_in_dataset(
)
trace_matrix_map = dataset_builder.trace_matrix_builder.trace_matrix_map
# TODO: clean up levels dependency
trace_dependency_graph = (dataset_builder.trace_matrix_builder.
create_trace_matrix_dependency_graph())
trace_ids: List[str] = trace_matrix_map.get_trace_ids()
paths = get_paths_to_complete_graph(trace_ids, trace_dependency_graph,
3)
self.assertEqual(1, len(paths.get_trace_ids()))
self.assertEqual(1, len(paths["0-2"]))
self.assertEqual([0, 1, 2], paths["0-2"][0])
def test_create_trace_matrices_in_definition(self):
dataset_builder = DatasetBuilder("MockDataset")
dataset_builder.structure_definition["traces"]["0-2"] = None
dataset_builder.artifact_builder.build()
dataset_builder.trace_matrix_builder.set_artifact_levels(
dataset_builder.artifact_builder.artifacts)
dataset_builder.trace_matrix_builder.build_matrices_defined_in_dataset(
)
trace_matrix_map_keys = (dataset_builder.trace_matrix_builder.
trace_matrix_map.get_trace_ids())
self.assertEqual(2, len(trace_matrix_map_keys))
self.assertIn("0-1", trace_matrix_map_keys)
self.assertIn("1-2", trace_matrix_map_keys)
self.assertNotIn("0-2", trace_matrix_map_keys)
self.assertEqual(0,
dataset_builder.trace_matrix_builder["0-1"].top_index)
self.assertEqual(
1, dataset_builder.trace_matrix_builder["0-1"].bottom_index)
def run(self) -> Table:
with create_loading_bar(
EXPERIMENT_NAME, DATASET_COLUMN_ORDER, len(DATASET_COLUMN_ORDER)
) as d_iterable:
for dataset_name in d_iterable:
builder = DatasetBuilder(dataset_name)
builder.build()
builder.export()
print(f"{dataset_name} exported.")
return Table()
def test_export_dataset(self):
"""
Tests that after MockDataset is exported all required folders have been updated.
:return:
"""
# Setup
dataset_name = "MockDataset"
dataset_builder = DatasetBuilder(dataset_name)
dataset_builder.build()
# Work
dataset_builder.export()
folders = ["Artifacts", "Oracles"]
for folder_rel_path in folders:
path_to_folder = os.path.join(
PATH_TO_SAMPLE_DATASETS, dataset_name, folder_rel_path
)
check_folder_has_updated(path_to_folder)
def test_get_paths_to_missing_traces_with_warc(self):
"""
Checks that the missing trace matrix in WARC is covered using a transitive path using the
defined trace matrices.
:return: None
"""
dataset_builder = DatasetBuilder("SAMPLE_WARC")
dataset_builder.artifact_builder.build()
dataset_builder.trace_matrix_builder.set_artifact_levels(
dataset_builder.artifact_builder.artifacts)
dataset_builder.trace_matrix_builder.build_matrices_defined_in_dataset(
)
graph = (dataset_builder.trace_matrix_builder.
create_trace_matrix_dependency_graph())
trace_ids: List[str] = list(
map(
to_string,
dataset_builder.trace_matrix_builder.trace_matrix_map.
get_trace_ids(),
))
paths = get_paths_to_complete_graph(trace_ids, graph, 3)
self.assertEqual(1, len(paths.get_trace_ids()))
self.assertEqual(1, len(paths["0-1"]))
self.assertEqual([0, 2, 1], paths["0-1"][0])
class TestTechniqueHelper(SmartTest):
d_name = "MockDataset"
d_builder = DatasetBuilder(d_name)
d_builder.build()
d_builder.export()
dataset = Dataset(d_name)
"""
Direct
"""
direct_algebraic_model = AlgebraicModel.VSM
direct_trace_type = TraceType.NOT_TRACED
direct_parameters = [direct_algebraic_model.value, direct_trace_type.value]
direct_components = ["0", "2"]
direct_definition = [
DIRECT_COMMAND_SYMBOL, direct_parameters, direct_components
]
"""
Intermediate
"""
transitive_algebraic_model = AlgebraicModel.VSM
transitive_aggregation_type = AggregationMethod.SUM
transitive_component_scaling_type = ScalingMethod.GLOBAL
transitive_component_trace_type = TraceType.NOT_TRACED
transitive_component_a = [
DIRECT_COMMAND_SYMBOL,
[
transitive_algebraic_model.value,
transitive_component_trace_type.value
],
["0", "1"],
]
transitive_upper_comp = "(%s (%s %s) (%s %s))" % (
DIRECT_COMMAND_SYMBOL,
transitive_algebraic_model.value,
transitive_component_trace_type.value,
"0",
"1",
)
transitive_component_b = [
DIRECT_COMMAND_SYMBOL,
[
transitive_algebraic_model.value,
transitive_component_trace_type.value
],
["1", "2"],
]
transitive_component_b_name = "(%s (%s %s) (%s %s))" % (
DIRECT_COMMAND_SYMBOL,
transitive_algebraic_model.value,
transitive_component_trace_type.value,
"1",
"2",
)
transitive_parameters = [
transitive_aggregation_type.value,
transitive_component_scaling_type.value,
]
transitive_components = [transitive_component_a, transitive_component_b]
transitive_technique_definition = [
TRANSITIVE_COMMAND_SYMBOL,
transitive_parameters,
transitive_components,
]
"""
Traced Components
"""
traced_component_type = TraceType.TRACED
traced_aggregation_value = AggregationMethod.MAX
traced_direct_component_a = [
DIRECT_COMMAND_SYMBOL,
[transitive_algebraic_model.value, traced_component_type.value],
["0", "1"],
]
traced_direct_component_b = [
DIRECT_COMMAND_SYMBOL,
[transitive_algebraic_model.value, traced_component_type.value],
["1", "2"],
]
traced_components = [traced_direct_component_a, traced_direct_component_b]
traced_parameters = [
traced_aggregation_value.value,
transitive_component_scaling_type.value,
]
"""
Sampled Artifacts
"""
sample_percentage = 0.5
sampled_parameters: [str
] = transitive_parameters + [repr(sample_percentage)]
sampled_components = transitive_components
sampled_artifacts_definition = [
SAMPLED_COMMAND_SYMBOL,
sampled_parameters,
sampled_components,
]
sampled_traces_definition = [
SAMPLED_TRACED_COMMAND_SYMBOL,
sampled_parameters,
sampled_components,
]
"""
Combined
"""
combined_aggregation_type = AggregationMethod.SUM
combined_parameters = ["SUM"]
combined_components = [direct_definition, transitive_technique_definition]
"""
Combined (with sampled transitive)
"""
combined_sampled_artifacts_components = [
direct_definition,
sampled_artifacts_definition,
]
combined_sampled_traces_components = [
direct_definition, sampled_traces_definition
]
direct_technique_name = "(. (VSM NT) (0 2))"
transitive_technique_name = (
"(x (SUM GLOBAL) ((. (VSM NT) (0 1)) (. (VSM NT) (1 2))))")
transitive_sampled_artifacts_technique_name = (
"(~ (SUM GLOBAL %f) ((. (VSM NT) (0 1)) (. (VSM NT) (1 2))))" %
sample_percentage)
transitive_sampled_traces_technique_name = (
"($ (SUM GLOBAL %f) ((. (VSM NT) (0 1)) (. (VSM NT) (1 2))))" %
sample_percentage)
combined_technique_name = "(o (%s) (%s %s))" % (
"SUM",
direct_technique_name,
transitive_technique_name,
)
combined_sampled_artifacts_technique_name = "(o (%s) (%s %s))" % (
"SUM",
direct_technique_name,
transitive_sampled_artifacts_technique_name,
)
combined_sampled_traces_technique_name = "(o (%s) (%s %s))" % (
"SUM",
direct_technique_name,
transitive_sampled_traces_technique_name,
)
def get_direct_definition(self) -> DirectTechniqueDefinition:
return DirectTechniqueDefinition(self.direct_parameters,
self.direct_components)
def get_transitive_definition(self) -> TransitiveTechniqueDefinition:
return TransitiveTechniqueDefinition(self.transitive_parameters,
self.transitive_components)
def get_traced_transitive_definition(
self) -> TransitiveTechniqueDefinition:
return TransitiveTechniqueDefinition(self.traced_parameters,
self.traced_components)
def get_combined_definition(self) -> HybridTechniqueDefinition:
return HybridTechniqueDefinition(self.combined_parameters,
self.combined_components)
def get_sampled_technique_definition(self) -> SampledTechniqueDefinition:
return SampledTechniqueDefinition(self.sampled_parameters,
self.sampled_components)
def get_combined_sampled_artifacts_definition(
self) -> HybridTechniqueDefinition:
return HybridTechniqueDefinition(
self.combined_parameters,
self.combined_sampled_artifacts_components)
def assert_valid_fake_dataset_similarity_matrix(
self, similarity_matrix: SimilarityMatrix):
self.assertEqual((1, 3), similarity_matrix.shape)
def create_counter_func(self, t_name: str):
n_function_calls = {"value": 0}
def counter_func(data: TechniqueData):
self.assertEqual(self.d_name, data.dataset.name)
self.assertEqual(t_name, data.technique.get_name())
n_function_calls["value"] = n_function_calls["value"] + 1
return counter_func, n_function_calls
def test_normalize_original_matrices(self):
db = DatasetBuilder("MockDataset")
db.build()
self.assertEqual(1, db.trace_matrix_builder["0-2"].matrix[0][2])
from api.datasets.builder.dataset_builder import DatasetBuilder
from api.datasets.dataset import Dataset
from api.tracer import Tracer
DATASET_NAME = "IllustrativeExample"
TOP_TECHNIQUE_NAME = "(. (VSM NT) (0 1))"
BOTTOM_TECHNIQUE_NAME = "(. (VSM NT) (1 2))"
DIRECT_TECHNIQUE_NAME = "(. (VSM NT) (0 2))"
TECHNIQUE_NAME = "(x (MAX INDEPENDENT) ((. (VSM NT) (0 1)) (. (VSM NT) (1 2))))"
REBUILD = False
if __name__ == "__main__":
if REBUILD:
dataset_builder = DatasetBuilder(DATASET_NAME)
dataset_builder.build()
dataset_builder.export()
dataset = Dataset(DATASET_NAME)
tracer = Tracer()
top_technique_data = tracer.get_technique_data(DATASET_NAME,
TOP_TECHNIQUE_NAME)
bottom_technique_data = tracer.get_technique_data(DATASET_NAME,
BOTTOM_TECHNIQUE_NAME)
direct_technique_data = tracer.get_technique_data(DATASET_NAME,
DIRECT_TECHNIQUE_NAME)
top_score = top_technique_data.similarity_matrix[0][0]
bottom_score = bottom_technique_data.similarity_matrix[0][0]
transitive_score = top_score * bottom_score
direct_score = direct_technique_data.similarity_matrix[0][0]