Friday, January 29, 2010

Information Structures


Information Structures

In an information system, data aggregation is necessary in order to ensure that the user has a clear overview of key interrelationships. Since the requirements which need to be satisfied by data aggregation are determined by many different user groups, it is absolutely vital that aggregation functionality is flexible. Experience shows that various user groups need to view aggregated data from many different perspectives.

The SAP system makes these different views possible by providing self-defined information structures.

In the information structures, you can define your own view, by selecting the information from the operative application that you consider worthy of aggregation.

A distinction is made between three types of information:

    You can define a maximum of nine characteristics for each information structure. These characteristics form the table keys of the database table.
  • Key figures
    The key figures refer to values with significance to business, which are aggregated by means of the characteristics (e.g. invoiced sales, quantity of incoming orders, order quantity, lead time).
  • Period split
    The periodic cumulation of data is a further criterion for aggregation. The key figures are thus aggregated on a daily, weekly or monthly basis, or on the basis of a variable period split that is determined by the fiscal year variant.

Information structures can be classified according to certain structure characteristics. The standard information structures have the same basic structure (for example, they all contain four period fields). Information structures with the same indicator have the same structure type.
The following types of information structures can be used in LIS:

  • Type '': Standard
    Information structure with periodic transaction data. This is the normal type of information structure.
  • Type 'C': Without period
    Information structure containing no period data. It contains current stock values, for example the indicator "Stock".
  • Type 'D': Without update
    This information structure contains no actual data (or planning data, as applicable). It is therefore not possible to create an update for this information structure. This structure is necessary to ensure the periodic display of stock values. The actual data can be found in an assigned information structure, type 'standard' and 'without period'. The corresponding standard analysis reads the data from the assigned information structures and carries out a back calculation of stock.
  • Type 'E': Standard (with stock values)
    Information structure which contains periodic key figures including periodic stock values. These stock values must be included in the information structure at the end of the period (or any point) using a batch program.
  • Type 'F': Document evaluation
    Information structure used to display document data. These information structures are subject to various performance measures due to the volume of data expected (for example, planning is not allowed). Additional functionality exists within the standard analysis (for example, total document analysis). Implementation area: Shopping Basket Analysis IS-R.
  • Type 'T': Transfer to SAP-BW
    Special information structure for transfers to the Business Information

Warehouse (BW). This information structure cannot be used in Reporting. These information structures are subject to performance measures (for example, planning not allowed) due to the volume of data expected.

Use of these different types is dependent on the application in which they are to be created:
Type ' ': All applications (default)
Type 'C': Application 01 (Sales & Distribution), 03 (Inventory Controlling) and 40 (Retail)
Type 'D': Application 03 (Inventory Controlling) and 40 (Retail)
Type 'E': Application 03 (Inventory Controlling) and 40 (Retail)
Type 'F': Application 40 (Retail)
Type 'T': All applications

When creating a customer-defined information structure using the attribute Planning possible, you must decide whether this is relevant or not. An information structure of type 'C' or type 'F' cannot be relevant to planning.

It is important to consider whether the information structure to be created is to be used in planning, as information structures that cannot be used in planning significantly increase the speed of the evaluation during reporting. Information structures not relevant to planning have their existing key fields in the table converted, so that non-relevant fields (SSOUR, period fields not required) can be positioned at the end of the key fields.

This speeds up the database search. If the self-defined information structure is to be populated with data, the primary index checks each table key. Version 000 is always used as the search criteria. The client is the first position in the database key, as otherwise each table in the data dictionary is interpreted as client-independent. The characteristics are positioned directly after the clients in the database key.

The sections of the database key appear as follows:

Client - characteristics - version - ...

When the update is activated, the database key is converted again, so that the period selected for the update appears earlier in the key.

Client - month - characteristics - version - ...

In order to avoid an accidental loss of data, key fields from an information structure not relevant to planning are only converted if there is no data in the accompanying database table.
It is important to note the following:

  • An information structure not relevant to planning can be used and updated with just one period (day, week, month, or posting period) in all clients of a system.
  • Changing a period results in the key fields being converted. It is only possible to change a period if none of the clients have any data in the database table anymore.

The self-defined and the standard information structures form the basis for all of the subsequent functions in the information system (e.g. analyses, planning).

No comments: