Forecast (Old)
Use
A forecast is a projection of existing data using statistical methods. You use forecast functions to draw conclusions from the historical development of key figures and to predict their likely future development.
Prerequisites
There must be data available that can serve as reference data for the forecast. In addition, the planning level, in whose context you create a forecast function, must contain at least one characteristic with time reference (for example, fiscal year).
Features
Based on existing plan or actual data, new plan data is generated for a certain time period. Various forecast strategies or models are supported and you determine these when you create a forecast function (you can make changes later if necessary). The available forecast strategies include the following:
· Exponential Smoothing (with and without Test for Trend and Seasonality)
· Test for Trend and Seasonality
To determine the data on which the forecast is to be based, specify the characteristics you want as reference data. The time characteristic you specified, which is required for each forecast function, is automatically included in the list of reference data.
You can determine that the forecast is only carried out for those data records that have a specific characteristic combination. You define these combinations by including the relevant characteristics as condition fields for the function. Apart from characteristics, you can also use the generic entry "key figure name" so that the function is carried out for certain key figures only. When you create a parameter group for the function, a column appears for each selected characteristic and you can enter the values here.
The time characteristic that you choose for a forecast function cannot be used at the same time in the same function as a characteristic for defining conditions. Therefore, the system automatically assigns this characteristic to the reference data and removes it from the list of characteristics for conditions.
See also: