SAP BW Data Model

1
SAP BW Data Model
2
Agenda

• SAP Namespace
• Basic Data Model
• Master Data
• Key Figure Additivity
• InfoProviders
• Slowly Changing Dimensions

3
Namespace

• SAP-delivery (Business Content) namespace
• Objects begin with 0
• Generated objects in the DDIC beginning with
  /BI0/
• InfoCube 0SALES, fact table /BI0/FSALES
• Customer namespace
• Objects beginning with A-Z
• Generated objects in the DDIC beginning with
  /BIC/
• InfoCube SALES, fact table /BIC/FSALES

4
Agenda

• SAP Namespace
• Basic Data Model
• Master Data
• Key Figure Additivity
• InfoProviders
• Slowly Changing Dimensions

5
Basic Approach

• SAP BI implements the star schema but has a
  different approach than most other data warehouse
  products
• Cannot create independent dimension or fact
  tables that can be reused
• Dimension and fact tables are created for each
  star schema
• Data reuse is achieved through the use of master
  data

6
InfoObjects

• InfoObjects are the basis of data modeling in SAP
  BW
• InfoObjects contain
• Key figures which are facts
• Characteristics which are attributes of
  dimensions and components of master data
• InfoObjects are contained in InfoObjectCatalogs
  which are grouped into InfoAreas
• There are separate InfoObjectCatalogs for key
  figures and for characteristics

7
InfoObject Organization
InfoAreas can be used to hierarchically organize
data into subject or application
areas. InfoObjectCatalogs appear at the bottom
of the hierarchy to contain key figures
andcharacteristics.
8
Characteristics
Characteristics

• Characteristics represent the fields of star
  schema dimensions
• However, they exist independently of dimension
  tables and can also be used as InfoProviders
• Characteristics are the unit of reusability in
  SAP BI rather than dimensions

9
Reference Characteristics

• A reference InfoObject can be specified when
  creating an InfoObject
• The newly created InfoObject inherits all the
  properties of the referenced object except
  textual descriptions and name, including its
  master data

10
Agenda

• SAP Namespace
• Basic Data Model
• Master Data
• Key Figure Additivity
• InfoProviders
• Slowly Changing Dimensions

11
Master Data

• Master data remains unchanged over a long period
  of time
• Characteristics in BI can bear master data
• Master data contains information that is needed
  again and again in the same way.
• For example, master data of supplier contains
  name, address, bank details
• One solution SAP offers to slowly changing
  dimensions
• Master data can be time dependent

12
Master Data

• Texts
• Text descriptions
• Actual name of person is assigned to attribute
  Person responsible for a cost center
• SAP way of replacing codes or difficult to
  interpret fields
• Texts are language dependent
• Attributes
• InfoObjects that are logically subordinate to a
  characteristic or should be permanently attached
  to the characteristic
• Person responsible for a cost center, customer
  address
• Hierarchies
• Characteristic values organized into a tree
  structure

13
SAP Master Data Model
14
Createsmastertexttables
Creates masterdata tables
Allows characteristicto be an InfoProvider
15
Slowly changingdimension is handled via
timeintervals.
16
Master Data Attributes arealsoCharacteristics
17
Hierarchies

• Characteristic hierarchy
• Characteristic values in a tree structure
• Can be used in all InfoProviders
• Example Cost center groups
• Dimensions of an InfoCube
• Example 0CALYEAR, 0CALMONTH, 0CALDAY
• Attributes of a characteristic
• Example 0MATERIAL, 0MATL_GROUP, 0MATL_TYPE, and
  0IND_SECTOR
• Must be navigation attributes

18
Characteristic Hierarchy in Query
19
Hierarchy in Dimension in Query
20
Dimension or Attribute Hierarchy Example
21
Characteristic Hierarchy Example
22
Characteristic Hierarchy Nodes

• Text nodes are folders which contain other nodes
• External Characteristic nodes contain a
  characteristic other than the one the hierarchy
  belongs to
• Value node contains a value of the characteristic
  which contains the characteristic hierarchy
• Postable nodes are those that represent a basic
  object of the hierarchy

23
Hierarchy Example
Text node
24
Unbalanced Hierarchy in BEx
25
Agenda

• SAP Namespace
• Basic Data Model
• Master Data
• Key Figure Additivity
• InfoProviders
• Slowly Changing Dimensions

26
Key Figures

• Key figures represent facts
• Key figures are required to be assigned a type
• Key figures of type amount are always assigned a
  fixed currency or an InfoObject that contains the
  currency must be assigned
• Key figures of type quantity must be assigned a
  fixed unit or must be assigned an InfoObject that
  contains the unit unit of measurement
• Key figures of type number, integer, data or time
  units of measure

27
Amount Key Figure
This Key Figure is of type Amount so either a
fixed currency must beassigned (as is here) or,
if the currency can vary, an InfoObject
thatwill contain the currency key mustbe
assigned.
28
Additivity of Key Figures

• Key figures have a aggregation function
  associated with them
• Additive key figures are modeled with the
  summation aggregation function
• Semiadditive key figures are assigned the
  summation aggregation function and are assigned
  an exception aggregation (such as maximum, first
  value or counter) for the nonadditive dimensions
• Nonadditive key figures are modeled as cumulative
  or noncumulative key figures with no aggregation
  function

29
Additivity of Key Figures
30
Agenda

• SAP Namespace
• Basic Data Model
• Master Data
• Key Figure Additivity
• InfoProviders
• Slowly Changing Dimensions

31
InfoProviders

• All reportable objects those that can be used
  with reporting tools are called InfoProviders
• Objects that physically contain data
• InfoCubes, Data Stores, InfoObjects that contain
  master data
• Logical views
• InfoSets, RemoteCubes, virtual InfoCubes and
  MultiProviders
• Dont physically contain data
• InfoProviders are grouped into InfoAreas

32
InfoProviders
InfoProviders are organized into InfoAreas. The
figure shows an InfoCube,Characteristics with
master data and a DataStore.
33
InfoCubes

• Implements the dimensional model
• Consist of a number of relational tables
• Time dimension uses builtin Time
  Characteristics
• Unit dimension containsconversion keys for
  currency and amountkey figures

34
Dimensions

• A dimension is a collection of characteristics
  whose content is logically connected under a
  single generic term
• While characteristics are available across all
  InfoCubes, dimensions represent a special
  organizational form within an InfoCube
• Cant create conformed dimensions that are
  available across the data warehouse

35
Degenerate Dimensions

• Degenerate dimensions are handled as properties
  of the InfoCube dimension
• The Line Item Dimension in the dialog below
  indicates that there is only one characteristic
  in the dimension
• No dimension tablewill be created

36
InfoCube Data Model
37
Agenda

• SAP Namespace
• Basic Data Model
• Master Data
• Key Figure Additivity
• InfoProviders
• Slowly Changing Dimensions

38
Versions of the Truth

• Historical truth
• Provides the status of entities at the time a
  transaction occurred and dont reflect changes in
  those entities that have occurred since
• Current truth
• Provides the status of entities as of today even
  if the status is different from what it was at
  the time of the transaction
• Truth at a specific point in time
• Provides the status of entities at a specific
  point in time regardless if the transaction
  occurred at a different time

39
Modeling Versions of the Truth

• Historical truth
• Modeled in InfoCubes
• Current truth
• Modeled as time independent navigational
  attributes or hierarchies
• Truth at a specific point in time
• Modeled as time dependent navigational attributes
  or hierarchies

40
Example Fact table stores revenues and product
group
41
Historical Truth
42
Example Fact table stores revenues
Product group is nowstored in an attribute of
the product characteristic
43
Current Truth
Using time independent master data
44
Truth as of 10/1/2005
Using time independent master data
45
Slowly Changing Dimensions

• Type 1
• Current truth
• Store current values with the facts
• Type 2
• Historical truth
• Store historic values with the facts
• Type 3
• Current truth and parts of historical truth
• Storing current values and one previous value