def test_lstm_model_fit_fail():
with pytest.raises(Exception):
model = lstm_model(**lstm_params)
x = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
y = np.array(
[100, 101, 102, 105, 107, 109, 110, 115, 119, 120, "random"])
model.fit(x, y)
def test_lstm_model_predict_fail():
with pytest.raises(Exception):
model = lstm_model(**lstm_params)
x = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
y = np.array([100, 101, 102, 105, 107, 109, 110, 115, 119, 120])
model.fit(x, y)
# wrong input type
result = model.predict([11])
def test_lstm_model_predict_pass():
model = lstm_model(**lstm_params)
x = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
y = np.array([100, 101, 102, 105, 107, 109, 110, 115, 119, 120])
model.fit(x, y)
result = np.round(model.predict(np.array([11])), 0)
expect = np.array([123])
assert np.allclose(result, expect)
def test_lstm_model_fit_pass():
model = lstm_model(**lstm_params)
x = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
y = np.array([100, 101, 102, 105, 107, 109, 110, 115, 119, 120])
model.fit(x, y)
result = model.lasty
expect = 120
assert result == expect
def update_prediction(self):
''' with new curren value and span, update the prediction
'''
try:
if self.model_type == "exp":
self.model = exponential_model(
**cattrs.param["prediction_models"]["exponential_model"])
if self.model_type == "lstm":
self.model = lstm_model(
**cattrs.param["prediction_models"]["lstm_model"])
if self.model_type == "logistic":
self.model = logistic_model(
**cattrs.param["prediction_models"]["logistic_model"])
# write model details
if self.model_type == "none":
self.pred_values = [0, 0, 0, 0, 0, 0, 0, 0]
self.model_details["model_name"] = "None"
self.model_details["r2"] = 0.0
self.model_details["msle"] = 0.0
self.model_details["show_prediction"] = False
else:
# update prediction values
x, y, x_test = np.array(self.keys_use), np.array(
self.values_use), np.array(self.pred_keys)
try:
self.model.fit(x, y)
self.pred_values = list(self.model.predict(x_test))
except: # if fail to fit or predict, use the last value as prediction
self.pred_values = list(
np.ones(x_test.shape) * self.values[-1])
# get last x day performance
try:
self.evaluate_model()
except: # if evaluation fails, show the failure on the web app
self.model_details = {
"r2": "Evaluation fails",
"msle": "Evaluation fails"
}
self.model_details["model_name"] = self.model.get_name()
self.model_details["show_prediction"] = True
except Exception as ex:
logger.error(Exception)
def test_lstm_model_get_name_fail():
with pytest.raises(Exception):
model = lstm_model(**lstm_params)
result = model.get_name("random")
def test_lstm_model_init_fail():
with pytest.raises(Exception):
# invalid input
model = lstm_model(nodes="random")
def test_lstm_model_get_name_pass():
model = lstm_model(**lstm_params)
result = model.get_name()
expected = "LSTM Model"
assert result == expected
def test_lstm_model_init_pass():
model = lstm_model(**lstm_params)
assert model.n_steps == 3
