Show TOC

Function documentationCalculating Time Series for the Artificial History

 

This function lets you build time series (artificial histories) for new products that are based on the time series of reference products. The artificial history consolidates the time series of the reference products – by forming the weighted mean value – to create a single time series, which SAP F&R uses for the new product in forecasting.

You can decide whether you want to create a new artificial history a the location product (which overwrites any existing time series) or update an existing artificial history.

Integration

  • You have to reconstruct the artificial history manually after the following events:

    • Change of the historical time series, DIFs, or listing data of references. For more information, see Editing Reference Data.

    • Change of the historical analysis period (the past period for the time series in the time series profile)

  • The system does not consider the following periods items when calculating the artificial history:

    • Periods that contain different DIF occurrences

    • Periods in which the reference product is not listed

    • Exclusion period

      Period in which the references do not have any logical time series values, for example, because the product does not yet have any even sales in the store or goods issues in the DC.

Features

  • To calculate the artificial history, you can schedule report /FRE/FU_HIST_REFMOD_CALC or start it manually.

    You can have the system analyze the status data to determine whether histories already exist, and update them if they do, or you can set the option to reconstruct the artificial history in any case.

    Once the system has reconstructed the artificial history, SAP F&R supplements it with additional time series values during each program run, until the valid-to date of the reference module is exceeded.

  • The system builds the history from the time series values of the references for each period as a weighted mean value.

    Example Example

    Calculation of the time series for a past period

    Three references with the weighting factors 1, 2 and 3 are assigned to a product.

    The consumption values 2 PC, 2 PC and 4 PC exist for the references in the analysis period.

    The weighted average is calculated as follows: (1 * 2 PC + 2 * 2 PC + 3 * 4 PC) / (1 + 2 + 3) = 18 PC / 6 = 3 PC.

    End of the example.

    Note Note

    If the system is unable to calculate the mean value for a period due to missing or excluded time series values, the time series for this period is not defined.

    End of the note.

  • The system checks and evaluates the following parameters (among others) when building the artificial history:

    • Start date for the time series

      • When you update the time series, SAP F&R calculates the artificial history from the Last History Setup date (saved in the status information).

      • When you reconstruct the time series, SAP F&R uses the number of periods that you defined in the forecast profile of the assigned references to calculate the start date of the time series.

    • Valid-to date, exclusion period, and listing period of the reference

    • Time series values of a reference product

      For a store, the system uses the time series for consumption data. For a distribution center, the system uses the time series that are also used for the forecast (these can be historical goods issues, aggregated historic store orders, and aggregated historic POS data).

  • The system takes factors for particular deviations of the time series values into account (DIFs). When a DIF is valid for all reference products, the system assigns it to the new product as well.

  • Depending on the DIF type and occurrence, the time series values of the reference products in the period concerned can be taken into account. For more information, see Consideration of Time Series Values Depending on DIF Type and Occurrence (Table).

  • The system saves the artificial history as the time series for the reference module. It passes a reconstructed history on to the F&R Processor, which takes it into account during the next forecast calculation, together with the level shift factor.

Activities

To start the setup of the artificial history manually, go to the SAP Easy Access screen and choose Start of the navigation path Forecasting and Replenishment Next navigation step Master Data Next navigation step Reference Module Next navigation step Calculate Artificial History End of the navigation path.

More Information

Time Series Aggregation