How Statistical Forecasts Are Disaggregated by Default

This topic explains how statistical forecasts are disaggregated by default in the Demand Management, Demand and Supply Planning, Planning Central, Replenishment Planning, and Sales and Operations Planning work areas.

Note: The default disaggregation described in this topic is limited to plans that use forecasting profiles that are based on Bayesian machine learning.

Settings That Affect Default Disaggregation of Statistical Forecasts

This table lists the forecasting parameters that affect the default disaggregation of statistical forecasts:

Note: In the Demand Management, Demand and Supply Planning, and Replenishment Planning work areas, you can modify these forecasting parameters in only user-defined forecasting profiles. You can use predefined forecasting profiles with only the default values.In the Planning Central and Sales and Operations Planning work areas, you can use only predefined forecasting profiles and can't modify forecasting parameters.

Name	Description	Default Value
PeriodsUntilActive	Specifies the number of days going backward from the history end date that's used for determining the forecast status for a product combination. If the demand for the product combination falls only within this period, the forecast status is Inactive/New/Immature. If the demand is wholly or partially before this period but within the days specified in the PeriodsUntilInactive forecasting parameter, the forecast status is Active. For example, if the value of the forecasting parameter is 14, but the only demand for the product combination was 10 days before the history end date, the forecast status is Inactive/New/Immature.	14 days
PeriodsUntilInactive	Specifies the number of days going backward from the history end date that's used for determining the forecast status for a product combination. If the demand for the product combination falls wholly before this period, the forecast status is Inactive/Retired/Discontinued. For example, if the value of the forecasting parameter is 182, but the only demand for the product combination was 200 days before the history end date, the forecast status is Inactive/Retired/Discontinued.	182 days
GlobalAllocationPeriods	Specifies the number of days going backward from the history end date that's used for calculating the average demand (for bookings, consumption, or shipments).	182 days

This table lists the forecast status measures and generate forecast measures for output measures of forecasting profiles:

Output Measure	Forecast Status Measure	Generate Forecast Measure
Bookings Forecast	Bookings Forecast Status Indicates if a product combination receives a disaggregated statistical forecast for bookings. Available in the Demand Management, Demand and Supply Planning, Planning Central, and Sales and Operations Planning work areas.	Bookings Forecast Generate Forecast Indicates whether historical bookings should be analyzed and a forecast generated for a product combination. You can edit this measure. Available in the Demand Management, Demand and Supply Planning, Planning Central, and Sales and Operations Planning work areas.
Consumption Forecast	Consumption Forecast Status Indicates if a product combination receives a disaggregated statistical forecast for consumption. Available in the Replenishment Planning work area.	Consumption Forecast Generation Status Indicates whether historical consumption should be analyzed and a forecast generated for a product combination. You can edit this measure. Available in the Replenishment Planning work area.
Shipments Forecast	Shipments Forecast Status Indicates if a product combination receives a disaggregated statistical forecast for shipments. Available in the Demand Management, Demand and Supply Planning, Planning Central, Replenishment Planning, and Sales and Operations Planning work areas.	Shipments Forecast Generate Forecast Indicates whether historical shipments should be analyzed and a forecast generated for a product combination. You can edit this measure. Available in the Demand Management, Demand and Supply Planning, Planning Central, Replenishment Planning, and Sales and Operations Planning work areas.
User-Defined Output Measure	<User-defined output measure name> Forecast Status (created automatically)	<User-defined output measure name> Generate Forecast Indicator (created automatically)

The table lists the history average measures for input measures of forecasting profiles:

Input Measure	History Average Measure
Final Bookings History	Bookings History Average Contains the average daily bookings over the days before the history end date as specified in the GlobalAllocationPeriods forecasting parameter. Available in the Demand Management, Demand and Supply Planning, Planning Central, and Sales and Operations Planning work areas.
Final Consumption History	Consumption History Average Contains the average daily consumption over the days before the history end date as specified in the GlobalAllocationPeriods forecasting parameter. Available in the Replenishment Planning work area.
Final Shipments History	Shipments History Average Contains the average daily shipments over the days before the history end date as specified in the GlobalAllocationPeriods forecasting parameter. Available in the Demand Management, Demand and Supply Planning, Planning Central, Replenishment Planning, and Sales and Operations Planning work areas.
User-Defined Input Measure	<User-defined input measure name> History Average (created automatically)

How Statistical Forecasts Are Disaggregated by Default

Whether a product combination receives a disaggregated statistical forecast depends on the forecast status of the combination, which can be one of the following:

Forecast Status	Description
1	Active A product combination has this forecast status if the demand is wholly or partially before the period specified in the PeriodsUntilActive forecasting parameter and wholly or partially within the period specified in the PeriodsUntilInactive forecasting parameter. Only product combinations with this forecast status receive a disaggregated forecast by default.
2	Inactive/New/Immature A product combination has this forecast status if the demand is wholly within the period specified in the PeriodsUntilActive forecasting parameter.
3	Inactive/Retired/Discontinued A product combination has this forecast status if the demand is wholly before the period specified in the PeriodsUntilInactive forecasting parameter.

These points explain the default disaggregation of statistical forecasts:

• The forecast status measure is identified on the basis of the output measure for the forecasting profile and calculated for a product combination after these criteria are considered:

  • The values of the PeriodsUntilActive and PeriodsUntilInactive forecasting parameters.

  • The value of the associated generate forecast measure for the product combination.

    Note: The generate forecast measure is an override for the forecast status measure. Edit the generate forecast measure for only those active product combinations that must not receive a forecast. Don't use this measure if you want inactive product combinations to receive a forecast.

• Depending on the input measure for the forecasting profile, the history average measure is selected. The history average for the product combination is then calculated on the basis of the GlobalAllocationPeriods forecasting parameter.

• The input history for all product combinations is collected irrespective of their forecast status (Active, Inactive/New/Immature, or Inactive/Retired/Discontinued) and aggregated to active product combinations at a higher level for generating the forecast.

• The generated forecast is disaggregated to only active product combinations at the lowest level on the basis of the history average (contained in the history average measure) and disaggregation parameters for the output measure of the forecasting profile.

As depicted in this figure, history is collected from active, inactive/new/immature, and inactive/retired/discontinued product combinations and aggregated to active product combinations at a higher level, and the forecast is disaggregated to only active product combinations at the lowest level.