Time Series Model Package
CFC Object
The CFC object creates optimized combinations of time series model forecasts. The performance of combined model forecasts tends to improve over that of individual models. You can specify the model forecasts to be combined either as arbitrary time series arrays or as instances of the TSM object that have been previously run. Table 5 summarizes the methods that are associated with the CFC object.
Table 5: Methods of the CFC Object
| Method | Description |
|---|---|
| AddModel | Add a forecast series from an instance of the TSM object |
| AddPredict | Add a time series array as a forecast series |
| criterion | Return the final fit statistic over a specified forecast region |
| GetForecast | Get the forecast series |
| Initialize | Initialize the CFC object |
| nfor | Return the forecast series length |
| Run | Run the CFC object to generate a combined forecast |
| SetOption | Specify the named option for the CFC object |
| SetY | Specify the dependent time series array (YSeries) for the CFC object |
The basic execution pattern for using a CFC object follows this sequence of operations:
Declare: The object declaration statement creates a new instance of the CFC object.
Initialize: The CFC.Initialize method resets the object in preparation for a new run.
SetY: The CFC.SetY method defines the dependent time series for the CFC object.
AddModel: The CFC.AddModel method adds the forecast series that was produced by an instance of the TSM object. Each call adds one model forecast series to the combination. Repeat as needed for each model forecast series to be combined. Also, if you invoke this method prior to invoking the CFC.SetY method, then this method will use the dependent series of the TSM object as the dependent series of the CFC object.
AddPredict: The CFC.AddPredict method adds a forecast series to the combination that is specified as a numeric array. Each call adds one forecast series to the combination. Repeat as needed for each forecast series to be combined.
SetOption: The CFC.SetOption method specifies any options that affect the generation of the combined model forecast. Each call defines an option. Repeat as needed to specify all options that are required.
Run: The CFC.Run method uses its currently configured dependent and forecast series data to generate a combined forecast.
Figure 2 illustrates the data flow through the CFC object and its relationship with other components of the TSM package.
Figure 2: CFC Object Data Flow
