You use conditions to structure a decision tree and to determine the individual steps for the demand combination run.
Conditions either compare the values of two key figures, or the values of one key figure with the constants within a certain period. Each comparison leads to either the result TRUE or FALSE. You assign each condition to its own node number. For both the TRUE result and the FALSE result, you specify the reference node number of a condition or of a finish activity that the system should subsequently evaluate.
In a condition you can compare the key figures of a specific planning area. If you use the comparison operator Equal, you can specify a factor for the threshold value with which you want the system to compare the key figure value. Otherwise the system checks for exact matches among the key figure values.
When the system compares key figure values over several periods, it determines a result for each period. Based on these results, it then determines a total result. If all the key figure values in each period fulfill the condition, the total result is TRUE. If one or more key figure values do not fulfill the condition, the total result is FALSE.
Depending on the result of the condition, the system can send an alert to the Alert Monitor.
You specify a processing horizon for each condition. This is a time frame within the demand combination horizon, in which the system is to compare the key figure values period by period. The processing horizon starts at the current period. You can also postpone the beginning of the processing horizon to a later period.
If you compare key figures that are based on different time horizons, for example from promotion planning, the comparison across individual periods might not lead to a result. For this type of comparison, the system can also cumulate the key figure values across all periods. For the system to do this, you define the cumulative horizon, in other words a rolling time span. The system determines the beginning of the cumulative horizon by calculating the time span backward from the start of the processing period. The system continually moves this time span by one period until the end of the processing period is reached.
The following two figures illustrate the processing horizon and the cumulative horizon, and the effects they have on the result.
The following figure uses calendar weeks (CW) as the periodicity. The current period is CW 40. The processing horizon begins two weeks after the current week, that is, CW 42. The processing horizon ends in CW 48. In the condition, the system is to compare the key figures 1 and 2 in this time frame and determine whether key figure 1 is greater than or equal to key figure 2.
The total result for this time frame is FALSE since the result does not correspond to the condition in CW 45.
In this figure, the processing horizon, the key figure values, and the condition are the same as in the previous figure. You have also defined a cumulative horizon of four periods. The cumulative horizon begins in CW 39.
During the demand combination run, the system first adds the key figure values for each data source for CW 39 to CW 42. It then adds the values for CW 40 to CW 44 and so on, until the end of the processing horizon is reached. The system then compares the individual sums for each data source and checks if the condition applies. Since the result in each period is TRUE, the total result is TRUE.