def arima():
"""
ARIMA (AutoRegressive Integrated Moving Average)
:return: list of prediction
"""
df = read_as_df("total.csv")
train_data, test_data = df[0:int(len(df) * 0.8)], df[int(len(df) * 0.8):]
df.head()
plot_train_and_test(train_data, test_data, df)
train_ar = train_data['price'].values
test_ar = test_data['price'].values
history = [x for x in train_ar]
print(type(history))
predictions = list()
for t in range(len(test_ar)):
model = ARIMA(history, order=(3, 1, 0))
model_fit = model.fit(disp=0)
output = model_fit.forecast()
yhat = output[0]
predictions.append(yhat)
obs = test_ar[t]
history.append(obs)
plot_prediction(df, test_data, predictions)
return predictions
def arima_validation(predictions):
df = read_as_df("total.csv")
train_data, test_data = df[0:int(len(df) * 0.8)], df[int(len(df) * 0.8):]
test_ar = test_data['price'].values
#error = mean_squared_error(test_ar, predictions)
#print('Testing Mean Squared Error: %.3f' % error)
#error2 = smape(test_ar, predictions)
#print('Symmetric mean absolute percentage error: %.3f' % error2)
plot_validation(df, test_data, predictions)
score = score_with_diff(test_ar.tolist(), predictions, 0.05, 0) # 5 cents
return score
def earning_calculation(pred_list, fee_p):
df = read_as_df("total.csv")
train_data, test_data = df[0:int(len(df) * 0.8)], df[int(len(df) * 0.8):]
train_arr = test_data['price'].values
test_list = test_data['price'].values.tolist()
# calculation
baseline = np.mean(train_arr)
inventory = 0
earning = 0
for i in range(len(pred_list)):
if pred_list[i] < baseline * (1 - fee_p) and inventory < 10:
inventory += 1
earning -= test_list[i] * (1 + fee_p)
elif pred_list[i] > baseline * (1 + fee_p) and 0 < inventory:
earning += test_list[i] * (1 - fee_p)
return earning + inventory * test_list[-1]
def earning_calculation(pred_list: [], fee_p: float) -> float:
df = read_as_df("total.csv")
train_data, test_data = df[0:int(len(df) * 0.8)], df[int(len(df) * 0.8):]
train_arr = test_data['price'].values
test_list = test_data['price'].values.tolist()
# calculation
baseline = np.mean(train_arr)
inventory = 0
earning = 0
total_val = 100000 # account initial values
for i in range(len(pred_list)):
if pred_list[i] < baseline * (1 - fee_p) and inventory < 10:
inventory += 1
earning -= test_list[i] * (1 + fee_p)
elif pred_list[i] > baseline * (1 + fee_p) and inventory > 0:
inventory -= 1
earning += test_list[i] * (1 - fee_p)
return earning + inventory * test_list[-1]
def test_write_file(self):
unit_test_file = "unit_test.csv"
write_as_new_file(unit_test_file, [["n1", 1], ["n2", 100]],
["col1", "col2"])
self.assertTrue(len(read_as_df(unit_test_file)) > 0)