def test_feature_vectors_calculator_cif(self):
"""
Test for cif file feature vectors calculator
"""
# logging test start
LOGGER.info('Feature vectors calculator test start')
# create MassTransit emulation dict
FEATURE_VECTORS_CALCULATOR_TEST['event_callback'] = feature_vectors_calculated
FEATURE_VECTORS_CALCULATOR_TEST['command_callback'] = feature_vectors_calculated
feature_vectors_calculated_publisher = MTPublisher(FEATURE_VECTORS_CALCULATOR_TEST)
# create valid message for send
message_body = {
'CorrelationId': 'c1cc1111-5d8b-0015-16ec-08d52f4261af',
'Fingerprints': [
{'Type': 'universal '}
],
'FileType': 'cif'
}
file_to_send = open(self.test_file_cif, 'rb')
REDIS_CLIENT.set(
'{}-file'.format(message_body['CorrelationId']),
file_to_send.read()
)
file_to_send.close()
# send message
fetch_token(self.oauth)
feature_vectors_calculated_publisher.publish(message_body)
time.sleep(10)
self.assertEqual(True, FEATURE_VECTORS_CALCULATED_FLAG)
def test_feature_vectors_calculator_failed(self):
"""
Test for feature vectors calculator failed case
"""
# logging test start
LOGGER.info('Feature vectors calculator failed case test start')
# create MassTransit emulation dict
FEATURE_VECTORS_CALCULATOR_FAIL_TEST['event_callback'] = feature_vectors_calculation_failed
FEATURE_VECTORS_CALCULATOR_FAIL_TEST['command_callback'] = feature_vectors_calculation_failed
feature_vectors_calculation_failed_publisher = MTPublisher(FEATURE_VECTORS_CALCULATOR_FAIL_TEST)
# create valid message for send
message_body = {
'CorrelationId': 'c1cc1111-5d8b-0015-16ec-08d52f4161fb',
'Fingerprints': [
{'Type': 'DESC'},
{'Type': 'FCFC', 'Size': 512, 'Radius': 3}
],
'FileType': 'sdf'
}
REDIS_CLIENT.set(
'{}-file'.format(message_body['CorrelationId']),
'empty'
)
# send message
fetch_token(self.oauth)
feature_vectors_calculation_failed_publisher.publish(message_body)
self.assertEqual(True, FEATURE_VECTORS_CALCULATION_FAILED_FLAG)
def test_properties_predicted_dnn(self):
self.test_file = 'dnn_model_dataset.h5'
dnn_dataset_blob_id = get_blob_id(self)
self.test_file = 'dnn_model.h5'
dnn_blob_id = get_blob_id(self)
LOGGER.info('Properties predicted test start')
PROPERTIES_PREDICTED_TEST[
'event_callback'] = properties_predicted
PROPERTIES_PREDICTED_TEST[
'command_callback'] = properties_predicted
message_body = {
'ParentId': self.parent_id,
'FolderId': self.user_id,
'FolderName': 'test_dnn.prediction',
'DatasetFileName': 'test_dnn',
'DatasetBlobId': dnn_dataset_blob_id,
'DatasetBucket': self.user_id,
'ModelBlobId': dnn_blob_id,
'ModelBucket': CLASSIC_CLASSIFICATION_MODEL_BUCKET,
'FingerprintRadius': 3,
'FingerprintSize': 1024,
'FingerprintType': 1,
'PrimaryIdField': 'Soluble',
'Id': 'd397a097-68eb-464a-86b2-abd2315daebf',
'UserId': self.user_id,
'CorrelationId': 'c1cc0000-5d8b-0015-16ec-08d52f4162fe'
}
properties_predicted_publisher = MTPublisher(
PROPERTIES_PREDICTED_TEST)
properties_predicted_publisher.publish(message_body)
def test_regression_properties_predicted(self):
"""
Test for 'regression classic properties successful predicted' workflow
"""
# logging test start
LOGGER.info('Regresion properties predicted test start')
# create MassTransit emulation dict
PROPERTIES_PREDICTED_TEST['event_callback'] = properties_predicted
PROPERTIES_PREDICTED_TEST['command_callback'] = properties_predicted
properties_predicted_publisher = MTPublisher(PROPERTIES_PREDICTED_TEST)
# create valid message for send
message_body = {
'ParentId': self.parent_id,
'FolderId': self.user_id,
'FolderName': 'test_regression_prediction',
'DatasetFileName': 'test',
'DatasetBlobId': self.source_blob_id,
'DatasetBucket': self.user_id,
'ModelBlobId': ELASTIC_NETWORK_MODEL_BLOB_ID,
'ModelBucket': CLASSIC_CLASSIFICATION_MODEL_BUCKET,
'Id': 'd397a097-68eb-464a-86b2-abd2315ddebf',
'UserId': self.user_id,
'CorrelationId': 'c1cc0000-5d8b-0015-16ec-08d52f4161fe'
}
# send message
fetch_token(self.oauth)
properties_predicted_publisher.publish(message_body)
self.assertEqual(True, CLASSIC_CLASSIFICATION_PREDICTED_FLAG)
def test_classic_classification_report_generation(self):
"""
Test for 'generate report' workflow
"""
# logging test start
LOGGER.info('Model trained test start')
# create MassTransit emulation dict
GENERATE_REPORT_TEST[
'event_callback'] = classic_classification_report_generation
GENERATE_REPORT_TEST[
'command_callback'] = classic_classification_report_generation
generate_report_publisher = MTPublisher(GENERATE_REPORT_TEST)
models_blob_ids = [
NAIVE_BAYES_MODEL_BLOB_ID,
LOGISTIC_REGRESSION_MODEL_BLOB_ID
]
# create valid message for send
message_body = {
'CorrelationId': 'c1cc0000-5d8b-0015-16ac-08d52f4161fe',
'ParentId': self.parent_id,
'TimeStamp': 'time stamp there',
'Models': create_models_data(models_blob_ids),
'UserId': self.user_id
}
# send message
fetch_token(self.oauth)
generate_report_publisher.publish(message_body)
self.assertEqual(True, True)
def test_classic_classification_model_training_naive_bayes(self):
"""
Test for 'model train' workflow
"""
# logging test start
LOGGER.info('Model trained test start')
# create MassTransit emulation dict
MODEL_TRAINED_TEST['event_callback'] = classic_classification_train_naive_bayes
MODEL_TRAINED_TEST['command_callback'] = classic_classification_train_naive_bayes
model_trained_publisher = MTPublisher(MODEL_TRAINED_TEST)
# create valid message for send
message_body = {
'SourceBlobId': self.source_blob_id,
'SourceBucket': self.user_id,
'ParentId': self.parent_id,
'Name': 'test_classic_classification_naive_bayes',
'Method': CODES[NAIVE_BAYES],
'ClassName': 'Soluble',
'SubSampleSize': 1.0,
'TestDatasetSize': 0.2,
'KFold': 5,
'Fingerprints': [
{'Type': 'DESC'},
{'Type': 'MACCS'},
{'Type': 'AVALON', 'Size': 512},
{'Type': 'FCFC', 'Size': 512, 'Radius': 3}
],
'CorrelationId': 'c1cc0000-5d8b-0015-2aab-08d52f3ea201',
'UserId': self.user_id,
'Scaler': 'Standard',
'HyperParameters': {
'optimizationMethod': 'default',
'numberOfIterations': 100
},
'DnnLayers': None,
'DnnNeurons': None
}
# send message
fetch_token(self.oauth)
model_trained_publisher.publish(message_body)
# wait while model training
time.sleep(10)
self.assertEqual(True, CLASSIC_CLASSIFICATION_NAIVE_TRAINED_FLAG)
def test_predictor_fail_event(self):
"""
Test for 'properties prediction failed' workflow
"""
# logging test start
LOGGER.info('Predictor fail test start')
# create MassTransit emulation dict
PREDICTOR_FAIL_TEST['event_callback'] = predictor_fail
PREDICTOR_FAIL_TEST['command_callback'] = predictor_fail
fail_predictor_publisher = MTPublisher(PREDICTOR_FAIL_TEST)
# create invalid message for send
message_body = {'123': 123}
# send message
fetch_token(self.oauth)
fail_predictor_publisher.publish(message_body)
self.assertEqual(True, PREDICTOR_FAIL_FLAG)
def test_modeler_fail_event(self):
"""
Test for 'modeling failed' workflow
"""
# logging test start
LOGGER.info('Modeler fail test start')
# create MassTransit emulation dict
MODELER_FAIL_TEST['event_callback'] = modeler_fail
MODELER_FAIL_TEST['command_callback'] = modeler_fail
fail_modeler_publisher = MTPublisher(MODELER_FAIL_TEST)
# create invalid message for send
message_body = {'123': 123}
# send message
fetch_token(self.oauth)
fail_modeler_publisher.publish(message_body)
self.assertEqual(True, MODELER_FAIL_FLAG)
def test_classic_classifier_single_properties_prediction(self):
"""
Test for classic regressor training optimizer
"""
# logging test start
LOGGER.info('Classic regression training optimizer test start')
# create MassTransit emulation dict
PREDICT_SINGLE_STRUCTURE_TEST[
'event_callback'] = classification_single_structure_predicted
PREDICT_SINGLE_STRUCTURE_TEST[
'command_callback'] = classification_single_structure_predicted
single_structure_property_predicted_publisher = MTPublisher(
PREDICT_SINGLE_STRUCTURE_TEST)
# create valid message for send
message_body = {
'CorrelationId': 'c1cc0000-5d8b-0015-16ec-08d52f4161fb',
'Id': 'd397a097-68eb-464a-86b2-abd2315ddeba',
'Structure': '\n Ketcher 5 71818102D 1 1.00000 0.00000 0\n\n 13 13 0 0 0 999 V2000\n -2.3818 -0.6252 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n -1.5159 -0.1254 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n -1.5157 0.8745 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0\n -0.6501 -0.6254 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0\n 0.2159 -0.1256 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n 0.2162 0.8747 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n 1.0824 1.3752 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n 1.9483 0.8753 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n 1.9479 -0.1251 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n 1.0818 -0.6256 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n 1.0820 -1.6254 0.0000 C 0 0 0 0 0 0 0 0 0 0 0 0\n 1.9479 -2.1252 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0\n 0.2161 -2.1254 0.0000 O 0 0 0 0 0 0 0 0 0 0 0 0\n 1 2 1 0 0 0 0\n 2 3 2 0 0 0 0\n 2 4 1 0 0 0 0\n 4 5 1 0 0 0 0\n 5 6 2 0 0 0 0\n 6 7 1 0 0 0 0\n 7 8 2 0 0 0 0\n 8 9 1 0 0 0 0\n 9 10 2 0 0 0 0\n 5 10 1 0 0 0 0\n 10 11 1 0 0 0 0\n 11 12 1 0 0 0 0\n 11 13 2 0 0 0 0\nM END\n',
'Format': 'MOL',
'PropertyName': 'Soluble',
'Models': [
{
'Id': str(uuid.uuid1()),
'Blob': {
'id': LOGISTIC_REGRESSION_MODEL_BLOB_ID,
'bucket': CLIENT_ID
}
},
{
'Id': str(uuid.uuid1()),
'Blob': {
'id': NAIVE_BAYES_MODEL_BLOB_ID,
'bucket': CLIENT_ID
}
}
]
}
# send message
fetch_token(self.oauth)
single_structure_property_predicted_publisher.publish(message_body)
self.assertEqual(True, CLASSIFICATION_SINGLE_STRUCTURE_PREDICTED_FLAG)
def test_regression_model_trained_event(self):
"""
Test for 'classic regression model successful trained' workflow
"""
# logging test start
LOGGER.info('Regression model trained test start')
# create MassTransit emulation dict
MODEL_TRAINED_TEST['event_callback'] = classic_regression_train
MODEL_TRAINED_TEST['command_callback'] = classic_regression_train
model_trained_publisher = MTPublisher(MODEL_TRAINED_TEST)
# create valid message for send
message_body = {
'CorrelationId': 'c1cc0000-5d8b-0015-2aab-08d52f3ea203',
'SourceBlobId': self.source_blob_id,
'SourceBucket': self.user_id,
'ParentId': self.parent_id,
'Name': 'test_regression',
'Method': CODES[ELASTIC_NETWORK],
'ClassName': 'logS',
'SubSampleSize': 1.0,
'TestDatasetSize': 0.2,
'KFold': 2,
'Fingerprints': [{'Type': 'ECFP', 'Radius': 4, 'Size': 512}],
'UserId': self.user_id,
'Scaler': 'Standard',
'HyperParameters': {
'OptimizationMethod': 'default',
'NumberOfIterations': 100
},
'DnnLayers': None,
'DnnNeurons': None
}
# send message
fetch_token(self.oauth)
model_trained_publisher.publish(message_body)
# wait while model training
time.sleep(10)
self.assertEqual(True, CLASSIC_REGRESSION_TRAINED_FLAG)
def test_dnn_classifying_model_trained_event(self):
"""
Test for 'DNN classifying model successful trained' workflow
"""
# logging test start
LOGGER.info('DNN classifying model trained test start')
# create MassTransit emulation dict
MODEL_TRAINED_TEST['event_callback'] = callback_models_trained
MODEL_TRAINED_TEST['command_callback'] = callback_models_trained
model_trained_publisher = MTPublisher(MODEL_TRAINED_TEST)
# create valid message for send
message_body = {
'SourceBlobId': self.source_blob_id,
'SourceBucket': self.user_id,
'ParentId': self.parent_id,
'Name': 'test_dnn_classifying',
'SourceFileName': 'test.sdf',
'Method': CODES[DNN_CLASSIFIER],
'ClassName': 'Soluble',
'SubSampleSize': 1.0,
'TestDatasetSize': 0.2,
'KFold': 2,
'Fingerprints': [{'Type': 'ECFP', 'Radius': 2, 'Size': 512}],
'CorrelationId': 'c1cc0000-5d8b-0015-2aab-08d52f3ea221',
'UserId': self.user_id,
'Scaler': 'Standard',
'HyperParameters': {
'optimizationMethod': 'default',
'numberOfIterations': 100
},
'DnnLayers': 2,
'DnnNeurons': 128
}
# send message
fetch_token(self.oauth)
model_trained_publisher.publish(message_body)
def test_training_optimizer_fail(self):
"""
Test for classic regressor training optimizer
"""
# logging test start
LOGGER.info('Training optimizer failed test start')
# create MassTransit emulation dict
OPTIMIZE_TRAINING_FAIL_TEST[
'event_callback'] = training_optimizer_fail
OPTIMIZE_TRAINING_FAIL_TEST[
'command_callback'] = training_optimizer_fail
properties_predicted_publisher = MTPublisher(
OPTIMIZE_TRAINING_FAIL_TEST)
# create valid message for send
message_body = {
'Methods': []
}
# send message
fetch_token(self.oauth)
properties_predicted_publisher.publish(message_body)
self.assertEqual(True, TRAINING_OPTIMIZER_FAIL_FLAG)
def test_classifier_training_optimizer(self):
"""
Test for classic classifier training optimizer
"""
# logging test start
LOGGER.info('Classic classification training optimizer test start')
# create MassTransit emulation dict
OPTIMIZE_TRAINING_TEST['event_callback'] = regression_training_optimized
OPTIMIZE_TRAINING_TEST['command_callback'] = regression_training_optimized
training_optimized_publisher = MTPublisher(OPTIMIZE_TRAINING_TEST)
# create valid message for send
message_body = {
'CorrelationId': 'c1ca0000-5d8b-0015-16ec-08d52f4161fe',
'Id': 'd397a097-68eb-464a-86b2-abd2315ddebf',
'UserId': self.user_id,
'SourceFileName': 'FileNameThere',
'TargetFolderId': self.parent_id,
'TimeStamp': '00',
'SourceBlobId': self.source_blob_id,
'SourceBucket': self.user_id,
'Methods': [CODES[NAIVE_BAYES], CODES[LOGISTIC_REGRESSION]],
'ClassName': 'Soluble',
'HyperParameters': {
'NumberOfIterations': 100,
'OptimizationMethod': 'default'
},
'DnnLayers': None,
'DnnNeurons': None
}
# send message
fetch_token(self.oauth)
training_optimized_publisher.publish(message_body)
self.assertEqual(True, REGRESSOR_TRAINING_OPTIMIZED)