import pandas as pd

# 读取数据
data = pd.read_csv('sales_data.csv', parse_dates=['date'])

# 查看数据前几行
print(data.head())

        date   sales
0 2019-01-01  1500.0
1 2019-01-02  1200.0
2 2019-01-03  1800.0
3 2019-01-04  1400.0
4 2019-01-05  1600.0

from statsmodels.tsa.stattools import adfuller

# 定义平稳性检验函数
def stationary_test(data):
    result = adfuller(data)

    print('ADF Statistic: %f' % result[0])
    print('p-value: %f' % result[1])

    if result[1] <= 0.05:
        print('序列是平稳的')
    else:
        print('序列是不平稳的')

# 提取销售额数据
sales = data['sales']

# 进行平稳性检验
stationary_test(sales)

ADF Statistic: -2.201718
p-value: 0.204952
序列是不平稳的

# 对销售额数据进行一阶差分
sales_diff = sales.diff().dropna()

# 进行平稳性检验
stationary_test(sales_diff)

ADF Statistic: -4.202571
p-value: 0.000631
序列是平稳的

from statsmodels.tsa.arima_model import ARMA

# 构建移动平均模型
model_ma = ARMA(sales_diff, order=(0, 1))
model_ma_fit = model_ma.fit(disp=0)

# 查看模型的AIC、BIC等指标
print(model_ma_fit.summary())

from statsmodels.tsa.arima_model import AR

# 构建自回归模型
model_ar = AR(sales_diff)
model_ar_fit = model_ar.fit()

# 查看模型的AIC等指标
print(model_ar_fit.summary())

from statsmodels.tsa.arima_model import ARMA

# 构建自回归移动平均模型
model_arma = ARMA(sales_diff, order=(2, 1))
model_arma_fit = model_arma.fit(disp=0)

# 查看模型的AIC、BIC等指标
print(model_arma_fit.summary())

from statsmodels.tsa.arima_model import ARIMA

# 构建差分自回归移动平均模型
model_arima = ARIMA(sales, order=(2, 1, 1))
model_arima_fit = model_arima.fit(disp=0)

# 查看模型的AIC、BIC等指标
print(model_arima_fit.summary())

# 进行预测
predictions = model_arima_fit.predict(start=len(sales), end=len(sales)+7, dynamic=False)

# 输出预测结果
print(predictions)