def test_prediction_endpoint_returns_prediction(flask_test_client):
# Load the test data from the classification_titanic package
# This is important as it makes it harder for the test
# data versions to get confused by not spreading it
# across packages
test_data = load_dataset(file_name=model_config.TESTING_DATA_FILE)
post_json = test_data[0:1].to_json(orient='records')
print(f"POST JSON {post_json}")
response = flask_test_client.post('/v1/predict/classification',
json=post_json)
assert response.status_code == 200
response_json = json.loads(response.data)
if isinstance(response_json['predictions'], list):
prediction = response_json['predictions'][0]
else:
prediction = response_json['predictions']
response_version = response_json['version']
assert isinstance(prediction, int)
assert response_version == _version
def test_make_single_prediction():
# Data
test_data = load_dataset(file_name=config.TESTING_DATA_FILE)
single_test_json = test_data[0:1].to_json(orient='records')
# Predict
subject = make_prediction(input_data=single_test_json)
# Test
assert subject is not None
assert isinstance(subject.get("predictions")[0], (int, np.integer))
def test_make_multiple_predictions():
# Data
test_data = load_dataset(file_name=config.TESTING_DATA_FILE)
original_data_length = len(test_data)
multiple_test_json = test_data.to_json(orient="records")
# Predict
subject = make_prediction(input_data=multiple_test_json)
# Test
assert subject is not None
# We manage missing values so we do not remove any row
assert len(subject.get('predictions')) == original_data_length
def run_training() -> None:
"""Train the model."""
# Read Training Data
data = load_dataset(file_name=config.TRAINING_DATA_FILE)
# Divide Train and Test sets (only if we read the original data set
X_train, X_test, y_train, y_test = train_test_split(data[config.FEATURES],
data[config.TARGET],
test_size=0.2,
random_state=0)
pipeline.titanic_pipe.fit(X_train, y_train)
_logger.info(f"saving model version: {_version}")
save_pipeline(pipeline_to_persist=pipeline.titanic_pipe)
def test_prediction_endpoint_validation_200(flask_test_client):
# Load the test data from the classification_titanic package
test_data = load_dataset(file_name=config.TESTING_DATA_FILE)
post_json = test_data.to_json(orient='records')
# Predict
response = flask_test_client.post('/v1/predict/classification',
json=post_json)
# Test
assert response.status_code == 200
response_json = json.loads(response.data)
# Check of correct number of errors removec if any
if response_json.get('errors') is None:
assert len(response_json.get('predictions')) == len(test_data)
else:
assert len(response_json.get('predictions')) + \
len(response_json.get('errors')) == len(test_data)