import numpy as np

# 你的數(shù)據(jù)：年份和對應(yīng)值
years = np.array([2023, 2024])
values = np.array([110, 100])

# 線性擬合：y = ax + b
a, b = np.polyfit(years, values, 1)

# 預(yù)測2025年
pred_2025 = a * 2025 + b
print(f"2025年預(yù)測值：{pred_2025:.1f}")
# 輸出：2025年預(yù)測值：90.0

from prophet import Prophet
import pandas as pd

# 假設(shè)你有月度數(shù)據(jù)
df = pd.DataFrame({
    'ds': pd.date_range('2023-01-01', periods=24, freq='MS'),  # 24個月
    'y':  [110, 108, 105, 103, 100, 98, 95, 97, 99, 102, 100, 98,
           96, 94, 91, 89, 88, 90, 92, 95, 93, 91, 89, 87]
})

model = Prophet(yearly_seasonality=True, weekly_seasonality=False)
model.fit(df)

future = model.make_future_dataframe(periods=12, freq='MS')  # 預(yù)測未來12個月
forecast = model.predict(future)

print(forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail(12))
model.plot(forecast)

import pandas as pd
import numpy as np

# 原始不規(guī)則數(shù)據(jù)
data = [
    ('2024-03-01', 100),
    ('2024-07-15', 90),
    ('2024-11-20', 85),
]
df = pd.DataFrame(data, columns=['date', 'value'])
df['date'] = pd.to_datetime(df['date'])

# 方法1：重采樣為季度（最簡單）
df_quarterly = df.set_index('date').resample('Q').mean()
print(df_quarterly)

# 方法2：線性插值補(bǔ)成月度
df_monthly = df.set_index('date').resample('MS').interpolate(method='linear')
print(df_monthly)

# 補(bǔ)完之后，就可以用Prophet或ARIMA了

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from statsmodels.tsa.stattools import adfuller
# 加載示例數(shù)據(jù)（如航空旅客數(shù)）
df = pd.read_csv('airline_passengers.csv', parse_dates=['Month'], index_col='Month')
series = df['Passengers']
# 平穩(wěn)性檢驗
result = adfuller(series)
print(f'ADF Statistic: {result[0]}, p-value: {result[1]}')  # p>0.05 非平穩(wěn)
# 一階差分
diff_series = series.diff().dropna()

from statsmodels.tsa.arima.model import ARIMA
from sklearn.metrics import mean_absolute_error

# 拆分訓(xùn)練/測試（前80%訓(xùn)練，后20%測試）
train = series[:int(0.8*len(series))]
test = series[int(0.8*len(series)):]

# 訓(xùn)練 ARIMA 模型 (示例參數(shù)，實際需調(diào)優(yōu))
model = ARIMA(train, order=(5,1,0))
fitted = model.fit()

# 預(yù)測
forecast = fitted.forecast(steps=len(test))
mae = mean_absolute_error(test, forecast)
print(f'MAE: {mae:.2f}')

from statsmodels.tsa.statespace.sarimax import SARIMAX

# 季節(jié)性周期 s=12（月度數(shù)據(jù)）
model = SARIMAX(train, order=(1,1,1), seasonal_order=(1,1,1,12))
fitted = model.fit()
forecast = fitted.forecast(steps=len(test))

from statsmodels.tsa.holtwinters import ExponentialSmoothing

model = ExponentialSmoothing(train, trend='add', seasonal='add', seasonal_periods=12)
fitted = model.fit()
forecast = fitted.forecast(len(test))

import xgboost as xgb
from sklearn.model_selection import train_test_split

# 構(gòu)造特征：使用過去 12 個值作為特征
def create_features(data, lookback=12):
    X, y = [], []
    for i in range(lookback, len(data)):
        X.append(data[i-lookback:i])
        y.append(data[i])
    return np.array(X), np.array(y)

values = series.values
X, y = create_features(values, lookback=12)

# 按時間順序劃分（不能打亂）
split = int(0.8 * len(X))
X_train, X_test = X[:split], X[split:]
y_train, y_test = y[:split], y[split:]

# 訓(xùn)練 XGBoost 模型
model = xgb.XGBRegressor(n_estimators=100, learning_rate=0.01)
model.fit(X_train, y_train)
y_pred = model.predict(X_test)

# 評估
mae = mean_absolute_error(y_test, y_pred)
print(f'XGBoost MAE: {mae:.2f}')

from sklearn.preprocessing import MinMaxScaler
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import LSTM, Dense

# 數(shù)據(jù)歸一化
scaler = MinMaxScaler()
scaled = scaler.fit_transform(series.values.reshape(-1,1))

# 構(gòu)造監(jiān)督學(xué)習(xí)樣本 (lookback=12)
lookback = 12
X, y = [], []
for i in range(lookback, len(scaled)):
    X.append(scaled[i-lookback:i, 0])
    y.append(scaled[i, 0])
X, y = np.array(X), np.array(y)

# 重塑為 LSTM 輸入格式 (樣本數(shù), 時間步長, 特征數(shù))
X = X.reshape((X.shape[0], X.shape[1], 1))

# 劃分訓(xùn)練/測試
split = int(0.8 * len(X))
X_train, X_test = X[:split], X[split:]
y_train, y_test = y[:split], y[split:]

# 構(gòu)建 LSTM 模型
model = Sequential()
model.add(LSTM(50, activation='relu', input_shape=(lookback, 1)))
model.add(Dense(1))
model.compile(optimizer='adam', loss='mse')

# 訓(xùn)練
model.fit(X_train, y_train, epochs=50, batch_size=32, verbose=0)

# 預(yù)測并反歸一化
y_pred_scaled = model.predict(X_test)
y_pred = scaler.inverse_transform(y_pred_scaled)
y_true = scaler.inverse_transform(y_test.reshape(-1,1))
mae = mean_absolute_error(y_true, y_pred)
print(f'LSTM MAE: {mae:.2f}')