Skip to content

Data mart settings

Data mart type

Aggregated data marts (With grouping)

  • Default type for new data marts
  • Groups records by included dimensions and base measures
  • Aggregated measures are calculated for each grouping
  • Can use fact table merging of type UNION or JOIN

Non-aggregated data marts (Without grouping)

  • Each record is displayed individually
  • Aggregated measures are calculated over the entire data set
  • Can only use fact table merging of type UNION

Changing the data mart type requires updating the data mart view in the database.

Merge type

UNION merging

  • Default merge type
  • Available for both data marts with grouping and without grouping
  • Combines rows from different fact tables
  • Automatically adds a "fact" field to identify the source fact table

JOIN merging

  • Available only for data marts with grouping
  • Does not add a "fact" field
  • Combines data based on common dimensions
  • Requires specific conditions:
    • No facts are added
    • Only dimensions from dimension groups are added
    • Only common dimension groups are used for all merged fact tables

Merge type selection:

  • Activated when the data mart contains elements from 2 or more fact tables
  • Can be changed at any time
  • Changing the merge type requires updating the data mart view in the database

Data mart filters

You can define filters for specific elements within a data mart to limit the included data.

Filterable elements:

  • Dimensions
  • Facts
  • Not available for measures

Filter configuration:

  • One filter per data mart element
  • The filter field is automatically selected based on the element
  • Configuration principle is the same as for metric verification filters
  • Option to apply a filter without including the filtered element in the view
  • Changing filters requires updating the data mart view in the database
  • Filters are saved with the data mart in the project version

Filter behavior:

  • Independent across different data marts
  • Copied during cloning and project version copying
  • Removed if the filtered element is deleted from the data mart
  • Applied during verification and SQL generation

Aggregation levels for measures

For measures added to data marts, you can define specific aggregation levels that differ from the general aggregation settings for the data mart.

Configuring aggregation level:

  • Performed by selecting one or more dimensions/facts already present in the data mart, over which aggregation is required
  • When selecting dimensions/facts for aggregation, the chosen aggregation level replaces the aggregation defined by the data mart type
    • For data marts without grouping:
      • If nothing is selected, no aggregation is performed
      • If some dimensions/facts are selected, aggregation is performed over those dimensions/facts
    • For data marts with grouping:
      • If nothing is selected, aggregation is performed over all dimensions/facts in the data mart
      • If some dimensions/facts are selected, aggregation is performed only over those dimensions/facts
  • Available for both data marts with grouping and without grouping
  • Aggregation levels can be changed at any time
  • Changing aggregation levels requires updating the data mart view in the database

Adding multiple instances of the same measure:

  • The same measure can be added multiple times with different aggregation levels
  • The system assigns unique names to distinguish instances based on the selected aggregation level
  • You can view how many times each measure has been added
  • If a measure is added multiple times with the same aggregation level, redundant instances are automatically removed when the data mart is saved