def add_sensor_component_feature(self, sliding_window: SlidingWindow,
sensor_component: SensorComponent,
feature: Feature, feature_df: DataFrame):
self.is_valid_data_set_manager()
training_data_set = self.workspace_repository.get_training_data_set(
self.workspace_id)
assert str(
sliding_window) in training_data_set.feature_extraction_cache
assert sensor_component not in training_data_set.feature_extraction_cache[str(
sliding_window
)].sensor_component_feature_df_file_IDs or feature not in training_data_set.feature_extraction_cache[
str(sliding_window
)].sensor_component_feature_df_file_IDs[sensor_component]
file_IDs_dict = training_data_set.feature_extraction_cache[str(
sliding_window)].sensor_component_feature_df_file_IDs
file_ID = self.file_repository.put_file(
FeatureExtractionData.serialize_sensor_component_feature_df(
feature_df))
if sensor_component in file_IDs_dict.keys():
file_IDs_dict[sensor_component][feature] = file_ID
else:
file_IDs_dict[sensor_component] = {feature: file_ID}
self.workspace_repository.set_training_data_set(
self.workspace_id, training_data_set)
def get_cached_sensor_component_feature(self,
sliding_window: SlidingWindow,
sensor_component: SensorComponent,
feature: Feature) -> DataFrame:
self.is_valid_data_set_manager()
training_data_set = self.workspace_repository.get_training_data_set(
self.workspace_id)
if str(
sliding_window
) not in training_data_set.feature_extraction_cache or sensor_component not in training_data_set.feature_extraction_cache[
str(
sliding_window
)].sensor_component_feature_df_file_IDs or feature not in training_data_set.feature_extraction_cache[
str(sliding_window)].sensor_component_feature_df_file_IDs[
sensor_component]:
raise RuntimeError(
"There is no cached feature extraction data with the given sliding window, sensor component and feature"
)
file_ID = training_data_set.feature_extraction_cache[str(
sliding_window
)].sensor_component_feature_df_file_IDs[sensor_component][feature]
return FeatureExtractionData.deserialize_sensor_component_feature_df(
self.file_repository.get_file(file_ID))
def get_feature_extraction_data_stub_5_1():
return FeatureExtractionData(
data_windows_df_file_ID=ObjectId("60707242b377f2b04ebf6737"),
labels_of_data_windows_file_ID=ObjectId("6070724cdd135f6e692c0959"),
sensor_component_feature_df_file_IDs={
"x_Accelerometer": {
Feature.MINIMUM: ObjectId("6070727ed1a8cbba14ea120f"),
Feature.MAXIMUM: ObjectId("6070728ce5eed4a716c39858")
},
"y_Accelerometer": {
Feature.MINIMUM: ObjectId("607072b8c5d9f62dc8af39c5"),
Feature.MAXIMUM: ObjectId("607072bd0cfbe6257fd2ccaa")
},
"z_Accelerometer": {
Feature.MINIMUM: ObjectId("607072c3c456e5966ed54877"),
Feature.MAXIMUM: ObjectId("607072c761d523b4852bc4fb")
},
"x_Gyroscope": {
Feature.MINIMUM: ObjectId("607072ccc9439e3d45c48a31"),
Feature.MAXIMUM: ObjectId("607072d1d7184b370db7a7f0")
},
"y_Gyroscope": {
Feature.MINIMUM: ObjectId("607072d5638d1b75b59e8990"),
Feature.MAXIMUM: ObjectId("607072dae852a74abbef32ca")
},
"z_Gyroscope": {
Feature.MINIMUM: ObjectId("607072dfcbc9bd19451f0be2"),
Feature.MAXIMUM: ObjectId("607072e37e4784db534231bf")
},
})
def get_feature_extraction_data_stub_4_2():
return FeatureExtractionData(
data_windows_df_file_ID=ObjectId("6070730112e067c7e6bf65df"),
labels_of_data_windows_file_ID=ObjectId("6070730644135e99d11e07e3"),
sensor_component_feature_df_file_IDs={
"x_Accelerometer": {
Feature.MEAN: ObjectId("6070730a3cb3407c3cee5088"),
Feature.MEDIAN: ObjectId("6070730f960d158336e60381")
},
"y_Accelerometer": {
Feature.MEAN: ObjectId("60707313d126b4fc5b5aad21"),
Feature.MEDIAN: ObjectId("6070731781da40c314d7fa59")
},
"z_Accelerometer": {
Feature.MEAN: ObjectId("6070731c86cd07184cb380ec"),
Feature.MEDIAN: ObjectId("607073213b96af39b7103af7")
},
"x_Gyroscope": {
Feature.MEAN: ObjectId("60707326750a110c0518ff57"),
Feature.MEDIAN: ObjectId("6070732a0464a34f7737b1c4")
},
"y_Gyroscope": {
Feature.MEAN: ObjectId("6070732e6c7d7545d2c3a6a5"),
Feature.MEDIAN: ObjectId("6070733273f8f9217b07af9d")
},
"z_Gyroscope": {
Feature.MEAN: ObjectId("6070733ab4862c8a4ea5591f"),
Feature.MEDIAN: ObjectId("6070733e37039f4f3f0a2bf1")
},
})
def add_split_to_windows(self, sliding_window: SlidingWindow,
data_windows: DataFrame,
labels_of_data_windows: List[str]):
self.is_valid_data_set_manager()
training_data_set = self.workspace_repository.get_training_data_set(
self.workspace_id)
assert str(
sliding_window) not in training_data_set.feature_extraction_cache
data_windows_df_file_ID = self.file_repository.put_file(
FeatureExtractionData.serialize_data_windows_df(data_windows))
labels_of_data_windows_file_ID = self.file_repository.put_file(
FeatureExtractionData.serialize_labels_of_data_windows(
labels_of_data_windows))
res = FeatureExtractionData(
data_windows_df_file_ID=data_windows_df_file_ID,
labels_of_data_windows_file_ID=labels_of_data_windows_file_ID)
training_data_set.feature_extraction_cache[str(sliding_window)] = res
self.workspace_repository.set_training_data_set(
self.workspace_id, training_data_set)
def get_cached_split_to_windows(
self, sliding_window: SlidingWindow) -> DataFrame:
self.is_valid_data_set_manager()
training_data_set = self.workspace_repository.get_training_data_set(
self.workspace_id)
if str(sliding_window
) not in training_data_set.feature_extraction_cache:
raise RuntimeError(
"There is no cached split to windows with the given sliding window"
)
file_ID = training_data_set.feature_extraction_cache[str(
sliding_window)].data_windows_df_file_ID
return FeatureExtractionData.deserialize_data_windows_df(
self.file_repository.get_file(file_ID))