def test_split_dataset():
df = train_model.load_dataset(train_model.DATA_FILE)
train_dataset, test_dataset = train_model.split_dataset(df)
assert isinstance(train_dataset, pandas.DataFrame)
assert isinstance(test_dataset, pandas.DataFrame)
assert train_dataset.shape == (26048, 15)
assert test_dataset.shape == (6513, 15)
def test_inference():
df = train_model.load_dataset(train_model.DATA_FILE)
train_dataset, test_dataset = train_model.split_dataset(df)
X_train, y_train, X_test, y_test, encoder, lb = \
train_model.process_datasets(train_dataset, test_dataset)
model = ml_model.load_model(train_model.MODEL_FILE)
preds = ml_model.inference(model, X_test)
assert isinstance(preds, numpy.ndarray)
assert preds.shape == (6513, )
def test_process_train_dataset():
df = train_model.load_dataset(train_model.DATA_FILE)
train_dataset, test_dataset = train_model.split_dataset(df)
X_train, y_train, encoder, lb = train_model.process_train_dataset(
train_dataset)
assert isinstance(X_train, numpy.ndarray)
assert isinstance(y_train, numpy.ndarray)
assert X_train.shape == (26048, 108)
assert y_train.shape == (26048, )
def test_compute_model_metrics():
df = train_model.load_dataset(train_model.DATA_FILE)
train_dataset, test_dataset = train_model.split_dataset(df)
X_train, y_train, X_test, y_test, encoder, lb = \
train_model.process_datasets(train_dataset, test_dataset)
model = ml_model.load_model(train_model.MODEL_FILE)
preds = ml_model.inference(model, X_test)
precision, recall, fbeta = ml_model.compute_model_metrics(y_test, preds)
assert isinstance(precision, numpy.float64)
assert isinstance(recall, numpy.float64)
assert isinstance(fbeta, numpy.float64)
def test_load_dataset():
df = train_model.load_dataset(train_model.DATA_FILE)
assert isinstance(df, pandas.DataFrame)
assert df.shape == (32561, 15)