def damped(data, forecast_length):
fitted_model = Holt(data, damped=True).fit()
prediction = fitted_model.predict(0, len(data) + forecast_length - 1)
params = fitted_model.params
params['initial_seasons'] = params['initial_seasons'].tolist()
return prediction, params
def MAPE(pred, org):
temp = np.abs((pred - org)) * 100 / org
return np.mean(temp)
#MAPE Mean Absolute Percentage Error (Lower is better)
#Subtract MAPE from 100 to provide accuracy in terms of percentage
# Simple Exponential Method
ses_model = SimpleExpSmoothing(Train["Passengers"]).fit()
pred_ses = ses_model.predict(start=Test.index[0], end=Test.index[-1])
se_MAPE = MAPE(pred_ses, Test.Passengers) # 7.846321
# Holt method
hw_model = Holt(Train["Passengers"]).fit()
pred_hw = hw_model.predict(start=Test.index[0], end=Test.index[-1])
hw_MAPE = MAPE(pred_hw, Test.Passengers) # 7.261176729658341
# Holts winter exponential smoothing with additive seasonality and additive trend
hwe_model_add_add = ExponentialSmoothing(Train["Passengers"],
seasonal="add",
trend="add",
seasonal_periods=12,
damped=True).fit()
pred_hwe_add_add = hwe_model_add_add.predict(start=Test.index[0],
end=Test.index[-1])
hwe_MAPE = MAPE(pred_hwe_add_add, Test.Passengers) # 4.500954
# Holts winter exponential smoothing with multiplicative seasonality and additive trend
hwe_model_mul_add = ExponentialSmoothing(Train["Passengers"],
seasonal="mul",
Train = sales.head(36)
Test = sales.tail(6)
model = []
mape = []
# Simple Exponential Smoothing Method
ses_model = SimpleExpSmoothing(Train["Sales"]).fit()
pred_ses = ses_model.predict(start=Test.index[0], end=Test.index[-1])
mape_ses = np.mean(np.abs((Test["Sales"] - pred_ses) / Test["Sales"])) * 100
model.append("Simple Exponential")
mape.append(mape_ses)
# Holts method / Double Exponential Smoothing Method
holt_model = Holt(Train["Sales"]).fit()
pred_holts = holt_model.predict(start=Test.index[0], end=Test.index[-1])
mape_holts = np.mean(np.abs(
(Test["Sales"] - pred_holts) / Test["Sales"])) * 100
model.append("Holts method / Double Exponential Smoothing Method")
mape.append(mape_holts)
# Holts Winter Exponential Smoothing with Additive Seasonality and Additive Trend
hwe_model_add_add = ExponentialSmoothing(Train["Sales"],
seasonal="additive",
trend="additive",
seasonal_periods=4,
damped=True).fit()
pred_hwe_add_add = hwe_model_add_add.predict(start=Test.index[0],
end=Test.index[-1])
mape_hwe_add_add = np.mean(