Restore Seasonality and Trend

This component restores seasonality (1st and 2nd) and trend into the forecasted residual series. The trend model, the seasonal components, and the lags where the seasonal peaks occur have been obtained in preprocessing the time series before training the forecasting model.

Input ports

1. Type: PMML
   Trend model
   Trend model as PMML
2. Type: Table
   Seed data
   Data containing columns for seasonal components. These were obtained in decomposing the time series training data.
3. Type: Table
   Forecasted data
   Forecasted residual series

Output ports

1. Type: Table
   Forecasted signal
   Forecasted signal values where the trend and seasonality components have been restored to the forecasted residuals