def test_validate_bag(self):
bag_generator = LoadMIL.bag_data_generator(pipeline='whole',
verbose=False)
bag, label = next(bag_generator)
self.assertTrue(LoadMIL.validate_bag(bag))
return None
def test_bag_data_generator(self):
bag_generator = LoadMIL.bag_data_generator(pipeline='whole',
verbose=False)
bag = next(bag_generator)
self.assertIsInstance(bag[0], csr_matrix)
return None
def setUp(self):
# Instance of BasePredictor
self.predictor = SVMCRBFMILESPredictor(
classifier_filename=SVMC_rbf_classifier_filename)
# Generate some data from dataclass / pydantic for FastAPI
self._generate_data()
# Load raw data from file or create raw data
(self.dfraw_load, self.bag_load,
self.bag_label_load) = LoadMIL.get_single_mil_bag(pipeline='whole')
return None
def setUp(self):
self.predictor = CompNBPredictor(
classifier_filename=CompNB_classifier_filename,
pipeline_type='categorical')
# Generate some data from dataclass / pydantic for FastAPI
self._generate_data()
# Load raw data from file or create raw data
(self.dfraw_load, self.bag_load,
self.bag_label_load) = LoadMIL.get_single_mil_bag(
pipeline='categorical')
return None
def setUp(self):
# Construct raw data input
# This is intended to test input gathered from a web form. Not all
# Attributes that are present in a SQL database are present
# Generate some data from dataclass / pydantic for FastAPI
self.input_data, self.input_data_list = generate_input_data()
self.input_data_pydantic, self.input_data_list_pydantic = generate_input_data_pydantic(
)
# Convert raw data to dataframe
self.dfraw_input = pd.DataFrame(data=[self.input_data.__dict__])
self.dfraw_input_pydantic = pd.DataFrame(
data=[self.input_data_pydantic.__dict__])
# Load raw data from file or create raw data
(self.dfraw_load, self.bag_load,
self.bag_label_load) = LoadMIL.get_single_mil_bag(pipeline='whole')
return None
from transform_mil import Transform
from dataclass_serving import RawInputData, RawInputDataPydantic
# Global declarations
config = configparser.ConfigParser()
config.read(r'../extract/sql_config.ini')
server_name = config['sql_server']['DEFAULT_SQL_SERVER_NAME']
driver_name = config['sql_server']['DEFAULT_SQL_DRIVER_NAME']
database_name = config['sql_server']['DEFAULT_DATABASE_NAME']
CompNB_classifier_filename = r"./compNB_si.clf"
KNN_classifier_filename = r"./knn_si.clf"
MultiNB_classifier_filename = r"./multiNB_si.clf"
SVMCL1SI_classifier_filename = r"./svmc_l1_si.clf"
SVMCRBFSI_classifier_filename = r"./svmc_rbf_si.clf"
LoadMIL = LoadMIL(server_name, driver_name, database_name)
#%% Class definitions
def generate_input_data():
# Construct raw data input
# This is intended to test input gathered from a web form. Not all
# Attributes that are present in a SQL database are present
input_data = RawInputData(
# Required numeric attributes
DEVICEHI=122.0,
DEVICELO=32.0,
SIGNALHI=10,
SIGNALLO=0,
SLOPE=1.2104,
if _PROJECT_DIR not in sys.path:
sys.path.insert(0, _PROJECT_DIR)
from mil_load import (load_mil_dataset_from_file, LoadMIL, bags_2_si,
bags_2_si_generator)
# Global declarations
config = configparser.ConfigParser()
config.read(r'../extract/sql_config.ini')
server_name = config['sql_server']['DEFAULT_SQL_SERVER_NAME']
driver_name = config['sql_server']['DEFAULT_SQL_DRIVER_NAME']
database_name = config['sql_server']['DEFAULT_DATABASE_NAME']
numeric_feature_file = config['sql_server']['DEFAULT_NUMERIC_FILE_NAME']
categorical_feature_file = config['sql_server'][
'DEFAULT_CATEGORICAL_FILE_NAME']
loadMIL = LoadMIL(server_name, driver_name, database_name)
#%%
def _filter_bags_by_size(X: Union[np.ndarray, csr_matrix, list],
min_instances: int, max_instances: int) -> np.ndarray:
"""Filter a set of bags by number of instances within the bag. If the
bag contains less than n_instances, then do not include that bag in the
returned index
inputs
-------
X: (np.ndarray or iterable) of bags
outputs
--------
index: (np.ndarray) index indicating where the number of instances per bag