Creating mappings

You create mappings to load structures to the hierarchies of an application model and to populate cubes with fact data. In both cases, mappings are from a source system, to the staging database, then to your application model. You can use tables and views as sources.

You create mappings on the Mappings tab of a cube or hierarchy in the Edit Database dashboard. If an existing mapping is similar to one that you want to create, you can duplicate the existing mapping and edit the duplicate as you require.

Mappings that you create are displayed in the Mappings Overview dashboard. There, you import, export, run, and delete mappings.

Cube mappings

Values for cube mappings may be held in a single column in the source table, or in multiple columns. For this reason, there are two value column modes: Single, and Dimension Driven.

Cube mappings with either single or dimension driven modes can be used for partial data loads. And the Process Latest Records option enables you to run incremental loads.

There are two cube mapping types: Source Column and Target Element. A single mapping can contain both types.

For the Source Column type, the element to be populated with data is read from the column in the source table that you specify.

For the Target Element type, the element to be populated is one that you specify. Or, you can use parameters as placeholders for target elements. This enables you to reuse a mapping with different parameter values. When you run the mapping, you are prompted to supply the required values.

Hierarchy mappings

For hierarchy mappings, the fields for each mapping depend on how the source data is structured. The source data can be held in one of these types of structure:

  • Level-based
  • Flat
  • Parent-child

These structures do not apply to fact data mappings, all of which are made through a data connection to a specified fact table.

The same data can be stored in different ways, so you must have knowledge of your source data. For example, a list of geographical areas could be stored in a table with a hierarchical, parent-child structure, or in a table with a level-based structure.

Level-based example

This table shows a level-based structure of areas. The three levels are Continent, Region, and Country.

Continent Region Country
Europe Central Slovakia
Europe Central Czech Republic
Europe Scandinavia Finland
Europe Scandinavia Sweden
United States Northeast Massachusetts
United States Northeast New York
United States Midwest Michigan
United States Midwest Minnesota

Flat structure example

An example of data with a flat structure is a list of currencies and their codes.

Parent-child example 1

This table is an example of parent-child relations, and the attribute values, in one table:

ID ParentID OrderPosition ISO_3166_2_CODE English German
Europe 1 Europe Europa
Central Europe Europe 1 Central Europe Zentraleuropa
Slovakia Central Europe 1 SK Slovakia Slowakei
Czech Republic Central Europe 2 CZ Czech Republic Tschechien
Scandinavia Europe 2 Scandinavia Skandinavien
Finland Scandinavia 3 FI Finland Finnland
Sweden Scandinavia 4 SE Sweden Schweden
United States 2 United States Vereinigte Staaten
Northeast United States 1 Northeast Nordost
Massachusetts Northeast 1 US-MA Massachusetts Massachusetts
New York Northeast 2 US-NY New York New York
Midwest United States 2 Midwest Midwest
Michigan Midwest 1 US-MI Michigan Michigan
Minnesota Midwest 2 US-MN Minnesota Minnesota
EU 3 European Union Europäische Union
Slovakia EU 3 SK Slovakia Slowakei
Czech Republic EU 2 CZ Czech Republic Tschechien
Finland EU 4 FI Finland Finnland
Sweden EU 1 SE Sweden Schweden

Parent-child example 2

This is an example of parent-child relations in separate tables.

This table contains the elements and the attribute values:

ID OrderPosition ISO_3166_2_CODE English German
Europe Europe Europa
Central Europe Central Europe Zentraleuropa
Slovakia SK Slovakia Slowakei
Czech Republic CZ Czech Republic Tschechien
Scandinavia Scandinavia Skandinavien
Finland FI Finland Finnland
Sweden SE Sweden Schweden
United States United States Vereinigte Staaten
Northeast Northeast Nordost
Massachusetts US-MA Massachusetts Massachusetts
New York US-NY New York New York
Midwest Midwest Midwest
Michigan US-MI Michigan Michigan
Minnesota US-MN Minnesota Minnesota
EU European Union Europäische Union

This table contains the relationships:

ID ParentID OrderPosition
Europe 1
Central Europe Europe 1
Slovakia Central Europe 1
Czech Republic Central Europe 2
Scandinavia Europe 2
Finland Scandinavia 3
Sweden Scandinavia 4
United States 2
Northeast United States 1
Massachusetts Northeast 1
New York Northeast 2
Midwest United States 2
Michigan Midwest 1
Minnesota Midwest 2
EU 3
Slovakia EU 3
Czech Republic EU 2
Finland EU 4
Sweden EU 1