The TD_KALMAN_FORECASTER function creates two layer ART tables that are ARTPRIMARY and ARTFITRESIDUALS. A primary result set containing point forecasts as a time series, the prediction intervals. The primary result is retrieved using SELECT on the analytical result table. Secondary layer ARTFITRESIDUALS containing residual values is retrieved using TD_EXTRACT_RESULTS.
| Name | Data Type | Description |
|---|---|---|
| ROW_I | Varies Data depends on OUTPUT_FMT(INDEX_STYLE()) specification. |
Row index of the series. |
| payloadN_FORECAST_VALUE | FLOAT | Point forecast values. |
| payloadN_LO_80 | FLOAT | Low end of the 80% prediction interval. |
| payloadN_HI_80 | FLOAT | High end of the 80% prediction interval. |
| payloadN_LO_95 | FLOAT | Low end of the 95% prediction interval. |
| payloadN_HI_95 | FLOAT | High end of the 95% prediction interval. |
These results can be retrieved by issuing a TD_EXTRACT_RESULTS against the ART using layer ARTFITRESIDUALS.
| Name | Data Type | Description |
|---|---|---|
| ROW_I | INTEGER | Row index associated with the residual series. |
| payloadNACTUAL_VALUE | FLOAT | The actual value from the input series. |
| payloadNCALC_VALUE | FLOAT | The calculated value of the response variable. |
| payloadNRESIDUAL | FLOAT | The difference between the calculated response value and the actual response value. |