You can improve system performance if you regularly run value statistics for classes with a large number of objects assigned to them. The value statistics can be used in decision making when setting up database accesses.
The system creates the statistics internally. All you need to do is call the function for creating value statistics.
The more frequently you create, change, and delete object assignments, the more frequently you should refresh your value statistics.
Structure of the Statistics
The characteristics are sorted according to the number of assigned objects with identical values for the characteristic. Characteristics that have different values assigned to different classified objects are more selective than characteristics that have the same value assigned to a large number of classified objects. The most selective characteristics are at the top of the statistics.
The statistics are based on the number of times a characteristic is used in a class type.
The statistics also provide information on the distribution of values. The value that has the least identical assignments is used to find objects.
Reference Characteristics
Reference characteristics are not normally included in value statistics, because the value for a reference characteristic is stored in a master data table, not in classification tables.
To create value statistics, choose in the class system menu
from the classification menu.Enter the class type and the client for which you want to create the statistics.
Choose Execute .