def test_holt_damp(self):
fit1 = SimpleExpSmoothing(self.livestock2_livestock).fit()
fit4 = Holt(self.livestock2_livestock,
damped=True).fit(damping_slope=0.98)
fit5 = Holt(self.livestock2_livestock, exponential=True,
damped=True).fit()
#We accept the below values as we getting a better SSE than text book
assert_almost_equal(fit1.params['smoothing_level'], 1.00, 2)
assert_almost_equal(fit1.params['smoothing_slope'], np.NaN, 2)
assert_almost_equal(fit1.params['damping_slope'], np.NaN, 2)
assert_almost_equal(fit1.params['initial_level'], 263.92, 2)
assert_almost_equal(fit1.params['initial_slope'], np.NaN, 2)
assert_almost_equal(fit1.sse, 6761.35, 2) #6080.26
assert_almost_equal(fit4.params['smoothing_level'], 0.98, 2)
assert_almost_equal(fit4.params['smoothing_slope'], 0.00, 2)
assert_almost_equal(fit4.params['damping_slope'], 0.98, 2)
assert_almost_equal(fit4.params['initial_level'], 257.36, 2)
assert_almost_equal(fit4.params['initial_slope'], 6.51, 2)
assert_almost_equal(fit4.sse, 6036.56, 2) #6080.26
assert_almost_equal(fit5.params['smoothing_level'], 0.97, 2)
assert_almost_equal(fit5.params['smoothing_slope'], 0.00, 2)
assert_almost_equal(fit5.params['damping_slope'], 0.98, 2)
assert_almost_equal(fit5.params['initial_level'], 258.95, 2)
assert_almost_equal(fit5.params['initial_slope'], 1.02, 2)
assert_almost_equal(fit5.sse, 6082.00, 2) #6100.11
def test_holt_damp(self):
fit1 = SimpleExpSmoothing(self.livestock2_livestock).fit()
mod4 = Holt(self.livestock2_livestock,damped=True)
fit4 = mod4.fit(damping_slope=0.98)
mod5 = Holt(self.livestock2_livestock,exponential=True,damped=True)
fit5 = mod5.fit()
#We accept the below values as we getting a better SSE than text book
assert_almost_equal(fit1.params['smoothing_level'],1.00, 2)
assert_almost_equal(fit1.params['smoothing_slope'],np.NaN, 2)
assert_almost_equal(fit1.params['damping_slope'],np.NaN, 2)
assert_almost_equal(fit1.params['initial_level'],263.92, 2)
assert_almost_equal(fit1.params['initial_slope'],np.NaN, 2)
assert_almost_equal(fit1.sse,6761.35, 2) #6080.26
assert_almost_equal(fit4.params['smoothing_level'],0.98, 2)
assert_almost_equal(fit4.params['smoothing_slope'],0.00, 2)
assert_almost_equal(fit4.params['damping_slope'],0.98, 2)
assert_almost_equal(fit4.params['initial_level'],257.36, 2)
assert_almost_equal(fit4.params['initial_slope'],6.51, 2)
assert_almost_equal(fit4.sse,6036.56, 2) #6080.26
assert_almost_equal(fit5.params['smoothing_level'],0.97, 2)
assert_almost_equal(fit5.params['smoothing_slope'],0.00, 2)
assert_almost_equal(fit5.params['damping_slope'],0.98, 2)
assert_almost_equal(fit5.params['initial_level'],258.95, 2)
assert_almost_equal(fit5.params['initial_slope'],1.02, 2)
assert_almost_equal(fit5.sse,6082.00, 2) #6100.11
def doHoltsLinear(train_set, test_set, predict_set):
print('>Holts Linear')
try:
# copy test dataframe dates
y_hat_avg = pd.DataFrame(index=test_set.index.copy())
# fit model
fit1 = Holt(np.asarray(train_set['Sales'])).fit(smoothing_level=0.3,
smoothing_slope=0.1)
# predict test dataframe
y_hat_avg['Sales'] = fit1.forecast(len(test_set))
# calculate error
rms = sqrt(mean_squared_error(test_set.Sales, y_hat_avg.Sales))
# create final predict dataframe
predict_set['FutureValue'] = fit1.forecast(len(predict_set))
# plot chart
#plotChart(train_set, test_set, y_hat_avg, 'Holt_linear', 'Sales')
except:
rms = 999999999
# return dataframes: error and prediction
return (rms, predict_set)
def __init__(self, data_in, num_prediction_periods):
self.__history = data_in
self.__num_prediction_periods = num_prediction_periods
y = np.array(data_in).reshape(-1, 1)
y = np.clip(y, 0.00001, None) # HoltWinters doesn't like zeros
self.__model = Holt(y, exponential=True, damped=True)
self.__results = self.__model.fit(optimized=True)
def Holt(self):
# model = Holt(self._train_data,damped=True)
with warnings.catch_warnings():
warnings.filterwarnings("ignore")
model = Holt(self._train_data.values)
model_fit = model.fit()
output = model_fit.forecast(self._h)
# print(output)
return output
def update(self):
begin = max(0, self.index - self.window)
data = self.arrivals[begin:self.index]
# fit model
model = Holt(data)
model_fit = model.fit()
self.model = model_fit
# make prediction
self.prediction = model_fit.predict(len(data), len(data))[0]
def holt(train, test, alpha=0.0, beta=0.0, exponential=False, damped=False):
if (alpha > 0.0 and beta > 0.0):
if (exponential or damped):
model = Holt(train, exponential=exponential,
damped=damped).fit(smoothing_level=alpha,
smoothing_slope=beta,
optimized=False)
else:
model = Holt(train).fit(smoothing_level=alpha,
smoothing_slope=beta,
optimized=False)
_alpha = '{0:2.1f}'.format(alpha)
_beta = '{0:2.1f}'.format(beta)
else:
if (exponential or damped):
model = Holt(train, exponential=exponential, damped=damped).fit()
else:
model = Holt(train).fit()
_alpha = model.model.params['smoothing_level']
_beta = model.model.params['smoothing_slope']
pred = model.predict(start=test.index[0], end=test.index[-1])
plt.plot(train.index, train, label='Train')
plt.plot(test.index, test, label='Test')
plt.plot(pred.index, pred, label=r'Holt, $\alpha={0}$'.format(_alpha))
plt.legend(loc='best')
def test_holt(self):
fit1 = Holt(self.air_ausair).fit(smoothing_level=0.8,
smoothing_slope=0.2,
optimized=False)
fit2 = Holt(self.air_ausair, exponential=True).fit(smoothing_level=0.8,
smoothing_slope=0.2,
optimized=False)
fit3 = Holt(self.air_ausair, damped=True).fit(smoothing_level=0.8,
smoothing_slope=0.2)
assert_almost_equal(fit1.forecast(5),
[43.76, 45.59, 47.43, 49.27, 51.10], 2)
assert_almost_equal(fit1.slope, [
3.617628, 3.59006512, 3.33438212, 3.23657639, 2.69263502,
2.46388914, 2.2229097, 1.95959226, 1.47054601, 1.3604894,
1.28045881, 1.20355193, 1.88267152, 2.09564416, 1.83655482
], 4)
assert_almost_equal(fit1.fittedfcast, [
21.8601, 22.032368, 25.48461872, 27.54058587, 30.28813356,
30.26106173, 31.58122149, 32.599234, 33.24223906, 32.26755382,
33.07776017, 33.95806605, 34.77708354, 40.05535303, 43.21586036,
43.75696849
], 4)
assert_almost_equal(fit2.forecast(5),
[44.60, 47.24, 50.04, 53.01, 56.15], 2)
assert_almost_equal(fit3.forecast(5),
[42.85, 43.81, 44.66, 45.41, 46.06], 2)
def update_graph(selected_dropdown1, selected_dropdown2):
df = values["Gold"]
trace1 = []
trace1.append(
go.Scatter(x=values.date,
y=df,
mode='lines',
opacity=0.6,
name='Actual values',
textposition='bottom center'))
if selected_dropdown1 != None and selected_dropdown2 != None:
alpha = float(selected_dropdown1)
beta = float(selected_dropdown2)
model = Holt(np.asarray(np.asarray(df)))
fit = model.fit(smoothing_level=alpha, smoothing_slope=beta)
result = fit.fittedvalues
name = 'Exponential Smoothing ' + '{0:.2f}'.format(
alpha) + ',' + '{0:.2f}'.format(beta)
trace1.append(
go.Scatter(x=values.date,
y=result,
mode='lines',
opacity=0.6,
name=name,
textposition='bottom center'))
title = ""
traces = [trace1]
data = [val for sublist in traces for val in sublist]
figure = {
'data':
data,
'layout':
go.Layout(colorway=[
"#5E0DAC", '#FF4F00', '#375CB1', '#FF7400', '#FFF400', '#FF0056'
],
height=600,
title=title,
xaxis={
"title": "Date",
'rangeslider': {
'visible': True
},
},
yaxis={"title": "Price (USD)"},
paper_bgcolor='rgba(0,0,0,0)',
plot_bgcolor='rgba(0,0,0,0)')
}
return figure
def daily_trend_pred(self, now, delta=pd.Timedelta(3, "W")):
train_time = pd.date_range(start=now - delta,
end=now - pd.Timedelta(1, "H"),
freq="H")
model_series = self.trend[train_time]
h_model = Holt(model_series, exponential=True, damped_trend=True)
#Fit
fitted_h = h_model.fit(optimized=True)
#Predict tomorrow
pred = fitted_h.predict(start=now, end=now + pd.Timedelta(23, "H"))
return pred
def hw(self, data, pre_len=7):
m = Holt(data).fit()
out = m.predict(pre_len)[-pre_len:].tolist()
self._get_model({
'model_name': 'holt_winter',
'model': m,
'pred': out,
'org_data': data,
'pre_len': pre_len
})
return out
def forecast(data, train_hours, test_hours):
# Train on the first 6 days to predict the 7th day
train = data.iloc[25:train_hours]
# Create the SES Model
model = Holt(np.asarray(train['seasonally differenced']), damped=True)
model._index = pd.to_datetime(train.index)
# Fit the model, and forecast
fit = model.fit()
pred = fit.forecast(test_hours)
data['holtDamped'] = 0
data['holtDamped'][25:] = list(fit.fittedvalues) + list(pred)
def forecast(data, train_hours, test_hours):
# Split data into training and test data
train = data.iloc[25:train_hours]
# Create the Holt Model
model = Holt(np.asarray(train['seasonally differenced']))
model._index = pd.to_datetime(train.index)
# Fit the model, and forecast
fit = model.fit()
pred = fit.forecast(test_hours)
data['holt'] = 0
data['holt'][25:] = list(fit.fittedvalues) + list(pred)
def trend(tr_dict):
"""
version of holt calculation, which fit and predict for each new sample
:param tr_dict:
:return:
"""
trends_dict = OrderedDict()
for k, v in tr_dict.items():
pred_des = [v[0], v[1]]
for i in range(2, len(v)):
des = Holt(v[:i]).fit(optimized=True)
pred_des.append(des.forecast(2)[0])
trends_dict[k] = np.array(pred_des)
return trends_dict
def halt(data):
""" halt(二次指数平滑)预测算法, 用于有明显趋势的数据 """
data_src=pd.Series(data)
#fit=Holt(data, damped=True).fit(smoothing_level=0.8, smoothing_slope=0.2)
fit=Holt(data, damped=False).fit(smoothing_level=0.8, smoothing_slope=0.2)
return fit
def Holtmethod(paramsList=['pollution.csv', '0.93','pm', 'humidity', 'date'], specialParams=['0.3','0.1']):
path = paramsList[0]
trainRows = float(paramsList[1])
saveto = 'result.csv'
df = pd.read_csv(path, usecols=paramsList[2:])
allRows = df.shape[0]
smoothing_level = specialParams[0]
smoothing_slope = specialParams[1]
train = df[0:int(allRows*trainRows)]
test = df[int(allRows*trainRows)+1:]
df['Timestamp'] = pd.to_datetime(df[paramsList[-1]], format='%Y/%m/%d %H:%M')
df.index = df['Timestamp']
df = df.resample('D').mean()
train['Timestamp'] = pd.to_datetime(train[paramsList[-1]], format='%Y/%m/%d %H:%M')
train.index = train['Timestamp']
train = train.resample('D').mean()
test['Timestamp'] = pd.to_datetime(test[paramsList[-1]], format='%Y/%m/%d %H:%M')
test.index = test['Timestamp']
test = test.resample('D').mean()
y_hat = test.copy()
nullArray = train.copy()
nullArray['time'] = train.index
# 以上可通用----------------------------
for i in range(2,len(paramsList)-1):
fit = Holt(np.asarray(train[paramsList[i]])).fit(smoothing_level=float(smoothing_level), smoothing_slope=float(smoothing_slope))
y_hat[paramsList[i]] = fit.forecast(len(test))
y_hat[paramsList[i]] = round(y_hat[paramsList[i]],2)
rms = sqrt(mean_squared_error(test[paramsList[i]], y_hat[paramsList[i]]))
print(rms)
y_hat['time'] = test.index
yhat_naive = np.array(y_hat)
nArray = np.array(nullArray)
newArray = np.concatenate((nArray,yhat_naive),axis=0)
s = pd.DataFrame(newArray, columns=paramsList[2:])
for i in range(2,len(paramsList)-1):
s[paramsList[i]][0:int(len(s)*trainRows)] = ""
s.to_csv(saveto,index=False,header=True,float_format='%.2f')
def pop_sim(init_data, num_increments):
data = init_data
for key, county in init_data.items():
population = pd.Series(county)
# https://www.statsmodels.org/stable/examples/notebooks/generated/exponential_smoothing.html#Holt's-Method
fit1 = Holt(np.asarray(population)).fit(smoothing_level=0.7,
smoothing_slope=0.3)
future_pop = fit1.forecast(num_increments)
last_inc = int(max(data[key].keys()))
for inc, value in zip(range(num_increments), future_pop):
# round negative population values to 0
data[key][str(last_inc + 1 + inc)] = value if value > 0 else 0
return data
def forecast(data, train_hours, test_hours, in_place=True):
train = data.iloc[25:train_hours]
model = Holt(np.asarray(train['seasonally adjusted']), damped=True)
fit = model.fit()
pred = fit.forecast(test_hours)
fcst = pd.concat(
[pd.Series([0] * 25), data['seasonal indices'][25:].multiply(
list(fit.fittedvalues) + list(pred)
)
]
)
if in_place:
data['holtDamped adjusted'] = fcst
else:
data['holtDamped adjusted'] = fcst
return fcst
def pipeline(attrs: dict, results: ResultContainer):
data = pd.read_csv(attrs["data_path"],
header=0,
index_col=0,
parse_dates=[0])
cpu_usage_data = data[["cpu_usage"]]
train_data_percentage = attrs["train_data_percentage"]
train_data_count = int((len(cpu_usage_data) * train_data_percentage) / 100)
train_data, test_data = cpu_usage_data[:train_data_count], cpu_usage_data[
train_data_count:]
model = Holt(train_data).fit(smoothing_level=attrs["smoothing_level"],
smoothing_slope=attrs["smoothing_slope"],
damping_slope=attrs["damping_slope"],
optimized=attrs["optimized"])
predictions = model.predict(start=train_data_count,
end=len(cpu_usage_data) - 1)
figure_train = figure(figsize=(10, 6))
plot(train_data[:200])
plot(model.fittedvalues[:200])
figure_predictions = figure(figsize=(10, 6))
plot(predictions[:200])
figure_result = figure(figsize=(10, 6))
plot(test_data[:200])
plot(predictions[:200])
mse = mean_squared_error(test_data, predictions)
rmse = sqrt(mse)
results.add_model_result("mse", mse)
results.add_model_result("rmse", rmse)
results.set_slack_main_figure(image_label="final-result",
figure=figure_result)
results.add_figure(image_label="train-plot", figure=figure_train)
results.add_figure(image_label="predictions-plot",
figure=figure_predictions)
def model(self, column_name, df, apply_smoothing, smoothing_level=None, smoothing_slope=None, damping_slope=None):
"""
performs predictions using the double exponential smoothing with damping model approach
:input column_name : str, name of column to hold the predicted values
:input df : dataframe, weekly-level data
:input apply_smoothing : bool, indicates whether to factor-in smoothing parameters in the Holt model
:input smoothing_level : int, default=None, l parameter in Holt model
:input smoothing_slope : int, default=None, b parameter in Holt model
:input damping_slope : int, default=None, phi parameter in Holt model
:returns df : dataframe, weekly-level, with predictions
:returns params : dictionary, default=None, placeholder for saving the best hyperparameters chosen by the model, if not provided as arguments to this method
"""
m = self.prediction_period
if apply_smoothing == True:
fit1 = Holt(df["train"][:-m],damped=True).fit(smoothing_level = smoothing_level, smoothing_slope = smoothing_slope,damping_slope = damping_slope, optimized = True)
params = None
elif apply_smoothing == False:
fit1 = Holt(df["train"][:-m],damped=True).fit(optimized = True)
params = fit1.params
if np.isnan(params['initial_slope']):
print('Initial Slope is Undefined')
fit1 = Holt(df["train"][:-m],damped=True).fit(damping_slope=0.1, optimized = True)
params = fit1.params
print('Model is refitted with damping slope fixed at 0.1')
if params['smoothing_slope']==0:
print('Smoothing Slope is 0')
fit1 = Holt(df["train"][:-m],damped=True).fit(smoothing_slope=0.1, optimized = True)
params = fit1.params
print('Model is refitted with smoothing slope fixed at 0.1')
print('====================')
print(params)
print('====================')
df[column_name] = np.nan
#y_fit = fit1.fittedvalues
y_fore = fit1.forecast(m)
#y_fore = fit1.predict(df.shape[0]-m)
#df[column_name][:-1] = y_fit
df[column_name][:-m] = df['train'].iloc[:-m]
df[column_name][-m:] = y_fore
return df
def valueForecast(file):
"""
电费预测
:param data: 电量数据
格式为：用户 日期 使用电量值
:return: 预测电量值
"""
logging.debug('开始运行')
data = pd.read_excel(file)
if data.shape[0] == 0:
raise ValueError('相关性原始数据不存在')
data.iloc[:, 0] = data.iloc[:, 0].astype(str)
users = set(data.iloc[:, 0].values)
# 用电量预测
result_pre = pd.DataFrame(columns=[
'DATA_DATE', 'DATA_DATE1', 'DATA_DATE2', 'DATA_DATE3', 'DATA_DATE4',
'DATA_DATE5'
])
for user in users:
subdata = data.loc[data.iloc[:, 0] == user]
df_index = pd.MultiIndex.from_frame(subdata.iloc[:, 1:2])
df = pd.DataFrame(np.array(subdata.iloc[:, -1]).reshape(1, -1),
columns=df_index)
df.dropna(axis=1, inplace=True)
df_values = df.values.flatten()
model = Holt(
endog=df_values,
initialization_method='estimated',
).fit()
pre = model.forecast(steps=5)
print(f'数据的预测 {pre}')
res2 = pd.DataFrame(pre).T
res2.columns = [
'DATA_DATE1', 'DATA_DATE2', 'DATA_DATE3', 'DATA_DATE4',
'DATA_DATE5'
]
res2['DATA_DATE'] = datetime.date.today()
res2['USRE'] = user
print(f'RES2 {res2}')
result_pre = result_pre.append(res2, ignore_index=True)
print(result_pre)
return result_pre
def get_trend(tr_dict, return_fitted=False):
"""
method of trend calculation, which fit holt
which returns also g_trend (i.e. the distance between y_holt - y_trajectory)
:param tr_dict:
:param return_fitted:
:return:
"""
trend_dict = OrderedDict()
res_dict = None
if return_fitted:
res_dict = OrderedDict()
for patient, values in tr_dict.items():
fitted = Holt(values, damped_trend=True).fit(optimized=True, smoothing_level=0.7) # type: HoltWintersResults
trend_dict[patient] = fitted.predict(start=0, end=len(values)-1)
if return_fitted:
res_dict[patient] = fitted.summary()
return trend_dict, res_dict
def double_exp_moving_average_forecast(self):
parameters = [[0.1, 0.3], [0.2, 0.8], [0.6, 0.6]]
self.create_fig_with_data()
for p, s in parameters:
fit1 = Holt(self.data_of_shares.Average).fit(smoothing_level=p,
smoothing_slope=s)
y_hat = fit1.fittedvalues
plt.plot(y_hat[0:self.amount_of_test_data],
label=f'Double exp smoothing level={p}, slope={s}')
plt.legend(loc='best')
plt.savefig("static/results/double_exp_moving_forecast")
def holt(alpha=0.2):
df = pd.read_csv('airline_passengers.csv',
index_col='Month',
parse_dates=True)
df['EWMA'] = df['Passengers'].ewm(alpha=alpha, adjust=False).mean()
df.index.freq = 'MS'
N_test = 12
train = df.iloc[:-N_test]
test = df.iloc[-N_test:]
train_idx = df.index <= train.index[-1]
test_idx = df.index > train.index[-1]
holt = Holt(df['Passengers'])
res_h = holt.fit()
df['Holt'] = res_h.fittedvalues
df[['Passengers', 'Holt']].plot()
plt.show()
holt = Holt(train['Passengers'])
res_h = holt.fit()
df.loc[train_idx, 'Holt'] = res_h.fittedvalues
df.loc[test_idx, 'Holt'] = res_h.forecast(N_test)
df[['Passengers', 'Holt']].plot()
plt.show()
class HoltWintersPredictor(object):
def __init__(self, data_in, num_prediction_periods):
self.__history = data_in
self.__num_prediction_periods = num_prediction_periods
y = np.array(data_in).reshape(-1, 1)
y = np.clip(y, 0.00001, None) # HoltWinters doesn't like zeros
self.__model = Holt(y, exponential=True, damped=True)
self.__results = self.__model.fit(optimized=True)
@property
def configuration(self):
return ""
def predict_counts(self):
start_index = len(self.__history)
y = self.__model.predict(self.__results.params,
start=start_index + 1,
end=start_index +
self.__num_prediction_periods)
y_list = y.ravel().tolist()
return clip(y_list, 0, inf)
def calculate_time_serie(data, time_serie_type, trend_seasonal, period,
forecast):
if time_serie_type == 'simpsmoothing':
data_simp_exp = SimpleExpSmoothing(data).fit()
proyeccion = data_simp_exp.forecast(int(forecast))
return data_simp_exp.fittedvalues, proyeccion
elif time_serie_type == 'holt':
data_holt = Holt(data).fit()
proyeccion = data_holt.forecast(int(forecast))
return data_holt.fittedvalues, proyeccion
elif time_serie_type == 'holt_winters':
print(trend_seasonal)
if trend_seasonal == 'add':
print('periodo', period)
data_holtwinters = ExponentialSmoothing(
data, trend='add', seasonal='add',
seasonal_periods=period).fit(use_boxcox=True)
print(data_holtwinters.fittedvalues)
elif trend_seasonal == 'mult':
data_holtwinters = ExponentialSmoothing(
data, trend='mul', seasonal='mul',
seasonal_periods=period).fit(use_boxcox=True)
proyeccion = data_holtwinters.forecast(int(forecast))
return data_holtwinters.fittedvalues, proyeccion
elif time_serie_type == 'arima':
arima = pmdarima.auto_arima(data,
seasonal=False,
error_action='ignore',
suppress_warnings=True)
proyeccion, int_conf = arima.predict(n_periods=int(forecast),
return_conf_int=True)
prediccion = arima.predict_in_sample()
print('pro', proyeccion)
print('pre', prediccion)
return prediccion, proyeccion
def predict_with_exp(self, dataset, code='PT', y='Total_Cases', days=5):
tmp = dataset[[y+code]]
history = [x for x in tmp.values]
news = []
data_mod = tmp.reset_index()
data_mod.columns = ["Date",y+code]
for i in range(days):
model = Holt(history[i:])
model_fit = model.fit(smoothing_level=self.level)
output = model_fit.forecast()
yhat = output[0]
history.append(np.array([round(yhat)]))
xn = datetime.datetime.strptime(data_mod.iloc[-1]['Date'], '%m/%d/%y') \
+ datetime.timedelta(days=i+1)
news.append(
pd.Series([xn.strftime("%m/%d/%y"), round(yhat)], index=data_mod.columns)
)
data_mod = pd.DataFrame(news)
data_mod.set_index('Date', inplace=True, drop=True)
data_mod.columns = ["EXPO"+code]
return data_mod
def holt_package(history_data):
result = Holt(history_data).fit(smoothing_level=0.93,
smoothing_slope=0.35,
optimized=False)
T = len(history_data)
forecast_data = [0] * T
PE = [0] * T
forecast_data = result.fittedvalues
forecast_data_last = result.fcastvalues
forecast_data = np.append(forecast_data, forecast_data_last)
for i in range(T):
PE[i] = abs(forecast_data[i] - history_data[i]) / history_data[i]
MPE = sum(PE) / T
print(result.summary)
return forecast_data_last, forecast_data, MPE
def __init__(self,
ts,
Mode,
Damped=False,
Trend=None,
Seasonal=None,
Seasonal_periods=None):
self.mode = Mode
self.ts = ts
if Mode == 'Simple':
self.exp_smoothing = SimpleExpSmoothing(ts)
elif Mode == 'Holt':
self.exp_smoothing = Holt(ts, damped=Damped)
elif Mode == 'Holt-Winter':
self.exp_smoothing = ExponentialSmoothing(
ts,
damped=Damped,
trend=Trend,
seasonal=Seasonal,
seasonal_periods=Seasonal_periods)
def double_exp_moving_average_errors(self):
parameters = [[0.1, 0.3], [0.2, 0.8], [0.6, 0.6]]
self.create_fig()
print("Double Exponential Average Method: ")
for p, s in parameters:
fit1 = Holt(self.data_of_shares.Average).fit(smoothing_level=p,
smoothing_slope=s)
y_hat = fit1.fittedvalues
mse = round(
mean_squared_error(self.test_data.Average,
y_hat[0:self.amount_of_test_data]), 6)
mae = round(
mean_absolute_error(self.test_data.Average,
y_hat[0:self.amount_of_test_data]), 6)
residual = self.test_data.Average - y_hat[0:self.
amount_of_test_data]
plt.plot(residual, label=f'Residual for level={p}, slope={s}')
print(f"MSE for level={p}, slope={s}: {mse}")
print(f"MAE for level={p}, slope={s}: {mae}")
plt.legend(loc='best')
plt.savefig("static/results/double_exp_moving_errors")
def exponential_smoothing(ts, label=None): """Exponential smoothing methods assign exponentially decreasing weights for past observations. """ from statsmodels.tsa.holtwinters import ExponentialSmoothing, SimpleExpSmoothing, Holt locs, labels = plt.xticks() model = SimpleExpSmoothing(np.asarray(ts)) # auto optimization fit1 = model.fit() pred1 = fit1.forecast(2) fit2 = model.fit(smoothing_level=.25) pred2 = fit2.forecast(2) fit3 = model.fit(smoothing_level=.5) pred3 = fit3.forecast(2) plot_exponential_smoothing(ts, fit1, fit2, fit3, pred1, pred2, pred3, label, "Simple Exponential Smoothing",figures_dir) model = ExponentialSmoothing(np.asarray(ts), trend='add') # auto optimization fit1 = model.fit() pred1 = fit1.forecast(2) fit2 = model.fit(smoothing_level=.9) pred2 = fit2.forecast(2) fit3 = model.fit(smoothing_level=.5) pred3 = fit3.forecast(2) plot_exponential_smoothing(ts, fit1, fit2, fit3, pred1, pred2, pred3, label, "Exponential Smoothing",figures_dir) model = Holt(np.asarray(ts)) # auto optimization fit1 = model.fit(smoothing_level=.3, smoothing_slope=.05) pred1 = fit1.forecast(2) fit2 = model.fit(optimized=True) pred2 = fit2.forecast(2) fit3 = model.fit(smoothing_level=.3, smoothing_slope=.2) pred3 = fit3.forecast(2) plot_exponential_smoothing(ts, fit1, fit2, fit3, pred1, pred2, pred3, label, "Holt Exponential Smoothing",figures_dir) return fit3
def test_holt(self):
fit1 = Holt(self.air_ausair).fit(smoothing_level=0.8,
smoothing_slope=0.2, optimized=False)
fit2 = Holt(self.air_ausair, exponential=True).fit(
smoothing_level=0.8, smoothing_slope=0.2,
optimized=False)
fit3 = Holt(self.air_ausair, damped=True).fit(smoothing_level=0.8,
smoothing_slope=0.2)
assert_almost_equal(fit1.forecast(5), [43.76,45.59,47.43,49.27,51.10], 2)
assert_almost_equal(fit1.slope,
[3.617628 ,3.59006512,3.33438212,3.23657639,2.69263502,
2.46388914,2.2229097 ,1.95959226,1.47054601,1.3604894 ,
1.28045881,1.20355193,1.88267152,2.09564416,1.83655482], 4)
assert_almost_equal(fit1.fittedfcast,
[21.8601 ,22.032368 ,25.48461872,27.54058587,
30.28813356,30.26106173,31.58122149,32.599234 ,
33.24223906,32.26755382,33.07776017,33.95806605,
34.77708354,40.05535303,43.21586036,43.75696849], 4)
assert_almost_equal(fit2.forecast(5),
[44.60,47.24,50.04,53.01,56.15], 2)
assert_almost_equal(fit3.forecast(5),
[42.85,43.81,44.66,45.41,46.06], 2)