Navigation :

Adding cube hierarchies

This page provides a description of how we can expose a datastore field as a cube level. It features an example in which we expose a calculated column ‘calculated’ from a new store as a level. For a full walk-through on how this store was added to the cube, see Adding New Data Loading or Unloading Topics.

tip

MR 3.1.0 introduced adding customizations through Spring Beans using new extension points. The example below assumes a new field has been added to the store, fed through the addition of a new column calculator. For documentation on the new mechanism, including how to implement such a column calculator, see Configuring sources.

The techniques employed are generic examples that can be extended, adapted and repeated for any use case that we need.

Step 1 - (Optional) Modify the schema selection description

If the default Atoti Market Risk behavior is not appropriate for the fields added in the datastore, schema selection customization might be required.

For details about the available customization mechanism, please refer to Configuring schema selections using Spring Beans.

Step 2 - Create the level

The field in the schema selection can now be turned into a level in the cube.

Dimensions, hierarchies, and levels can be added to the cube configuration by creating Spring Beans with type HierarchyBuilderConsumer and one of the following qualifiers:

Cube	Qualifier
Var-ES Cube	varDimensions
Sensitivity Cube	sensiDimensions, sensiCommonDimensions
PLCube	aPnlDimension
Var-ES Summary Cube	varSummaryDimensions
Sensitivity Summary Cube	sensiSummaryDimensions
PL Summary Cube	aPnlSummaryDimension
Market Data Cube	mdDimensions

For example:

    @Qualifier(SP_QUALIFIER__VAR_DIMENSIONS)
    @Order(50)
    @Bean
    public HierarchyBuilderConsumer myHierarchy() {
        return hierarchyBuilder -> hierarchyBuilder.toDimension("Sign", builder -> builder
            .withHierarchy("Sign")
            .withLevel("calculated");
    }

Note: The order has been set to 50 so that this dimension is added after the existing dimensions (order=20) and before the epoch dimension (order=99).

When using DirectQuery

When using DirectQuery, perform the steps as listed preceding. There are no additional requirements.

warning

The DirectQuery and in-memory Atoti servers must be configured identically. There must not be any hierarchies in the DirectQuery data node that do not exist in the in-memory one.

Copper hierarchies case

The Copper hierarchies are analysis hierarchies that are built from Copper join from a side store. You can also amend Copper hierarchies. They mainly work the same way as the Copper metrics.

The Copper join

The copper join is exposed on a bean of type CopperStore, with a qualifier. You can override it by creating a new bean with the same qualifier and the primary annotation. For instance if you want to override the Var-ES cube to Scenario store Copper join:

    @Primary
    @Qualifier(SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_VAR)
    @Bean
    public CopperStore myScenarioStoreForVaR() {
        return Copper.store(SCENARIO_STORE_NAME)
           .joinToCube(UnlinkedCopperStore.JoinType.LEFT)
           [...];
    }

The Copper hierarchy

The Copper hierarchies are exposed on beans of type Publishable<CopperHierarchy>, with a qualifier. You can override it by creating a new bean with the same qualifier and the primary annotation. For instance if you want to override the Var-ES cube liquidity Horizon Copper hierarchy:

    @Primary
    @VarCopperContextBean
    @Qualifier(SP_QUALIFIER__LIQUIDITY_HORIZONS_HIERARCHY_VAR)
    public Publishable<CopperHierarchy> liquidityHorizonHierarchyVaR(@Qualifier(SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_VAR) CopperStore scenarioStoreForVaR) {
        return Copper
                .newHierarchy(RISK_DIMENSION, LIQUIDITY_HORIZONS_HIERARCHY)
                .fromStore(scenarioStore)
                .withLevel(LIQUIDITY_HORIZON_LEVEL, FieldPath.of(StoreFieldConstants.LIQUIDITY_HORIZON));
    }

You can also add a new Copper hierarchy by creating a new bean of type Publishable<CopperHierarchy> with a new qualifier.

The Copper hierarchy qualifiers

The following table lists the Copper hierarchy qualifiers that are used in the different cubes. You can override any of them by creating a new bean with the same qualifier and the primary annotation.

Store name	Cube name	Hierarchy	Qualifier	CopperStore qualifier
Scenarios	Sensitivity Cube	Scenarios@Risk	SP_QUALIFIER__SCENARIO_ANALYSIS_HIERARCHY_SENSI	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_SENSI
Scenarios	Sensitivity Summary Cube	Scenarios@Risk	SP_QUALIFIER__SCENARIO_ANALYSIS_HIERARCHY_SENSI_SUMMARY	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_SENSI_SUMMARY
Scenarios	Var-ES Cube	Scenarios@Risk	SP_QUALIFIER__SCENARIO_ANALYSIS_HIERARCHY_VAR	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_VAR
Scenarios	Var-ES Summary Cube	Scenarios@Risk	SP_QUALIFIER__SCENARIO_ANALYSIS_HIERARCHY_VAR_SUMMARY	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_VAR_SUMMARY
Scenarios	Sensitivity Cube	Liquidity Horizons@Risk	SP_QUALIFIER__LIQUIDITY_HORIZONS_SENSI	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_SENSI
Scenarios	Sensitivity Summary Cube	Liquidity Horizons@Risk	SP_QUALIFIER__LIQUIDITY_HORIZONS_SENSI_SUMMARY	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_SENSI_SUMMARY
Scenarios	Var-ES Cube	Liquidity Horizons@Risk	SP_QUALIFIER__LIQUIDITY_HORIZONS_HIERARCHY_VAR	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_VAR
Scenarios	Var-ES Summary Cube	Liquidity Horizons@Risk	SP_QUALIFIER__LIQUIDITY_HORIZONS_HIERARCHY_VAR_SUMMARY	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_VAR_SUMMARY
Scenarios	Sensitivity Cube	Scenario Sets@Risk	SP_QUALIFIER__SCENARIO_SET_HIERARCHY_SENSI	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_SENSI
Scenarios	Sensitivity Summary Cube	Scenario Sets@Risk	SP_QUALIFIER__SCENARIO_SET_HIERARCHY_SENSI_SUMMARY	SP_QUALIFIER__SCENARIO_STORE_NAME_JOIN_SENSI_SUMMARY

The left-join Copper metrics

The left-join Copper metrics are handled on the same exact way as the other Copper metrics are. The only difference is that you need to inject the qualified CopperStore bean that you want to use for the left-join.