从固定价值到原始价值

Question

import statsmodels.api as sm
sarimax_mod6 = sm.tsa.statespace.SARIMAX(endog = train_sarima_df['QtySupplied'][:end_index],
                                        exog = exog_data[:end_index],  
                                        trend='n', order=(6,1,2), seasonal_order=(0,1,1,7)).fit()

sarima_mod6.predict(start = start_index, end= end_index, dynamic= True)  
sarimax_mod6.forecast(steps = 31,exog = exog_data['2019-9-1':'2019-9-30'])

对于代码来说，预测结果看起来已经变得平稳了。有没有办法获得真实的预测值？