def save_model(self, config, label_performance_metrics, column_order,
label_encoder, pipeline):
self.is_valid_data_set_manager()
label_encoder_object_file_ID = self.file_repository.put_file(
MlModel.serialize_label_encoder(label_encoder))
pipeline_object_file_ID = self.file_repository.put_file(
MlModel.serialize_pipeline(pipeline))
model_in_db = MlModel(None, config, label_performance_metrics,
column_order, label_encoder_object_file_ID,
pipeline_object_file_ID)
model_id = self.ml_model_repository.add_ml_model(model_in_db)
self.workspace_repository.add_ml_model_ref(self.workspace_id, model_id)
def get_ml_model_stub_4_2() -> MlModel:
return MlModel(
_id="607367362d98418cae5a1522",
config=get_training_config_stub_4_2(),
label_performance_metrics=get_label_performance_metrics_stub_4_2(),
column_order=get_column_order_stub_4_2(),
label_encoder_object_file_ID="60736744c795e80494dbc3e9",
pipeline_object_file_ID="60736749a4c260179e9be00d")
def get_ml_model_stub_5_1() -> MlModel:
return MlModel(
_id="60736013b961d239c76711a3",
config=get_training_config_stub_5_1(),
label_performance_metrics=get_label_performance_metrics_stub_5_1(),
column_order=get_column_order_stub_5_1(),
label_encoder_object_file_ID="6073604f8741014a0cd09780",
pipeline_object_file_ID="607360560b3beb16ac03de77")
def add_ml_model(self, ml_model: MlModel) -> ObjectId:
id = ObjectId(random_object_id.generate())
ml_model._id = id
self.models[id] = ml_model
return id
def get_pipeline(self) -> Pipeline:
ml_model = self.get_ml_model()
return MlModel.deserialize_pipeline(
self.file_repository.get_file(ml_model.pipeline_object_file_ID))
def get_label_encoder(self) -> LabelEncoder:
ml_model = self.get_ml_model()
return MlModel.deserialize_label_encoder(
self.file_repository.get_file(
ml_model.label_encoder_object_file_ID))
def add_ml_model(self, ml_model: MlModel) -> ObjectId:
ml_model_dict = ml_model.dict_for_db_insertion()
return self.collection.insert_one(ml_model_dict).inserted_id