Daanse Tutorial - Dimension Optimisation With Level Attribute
A basic OLAP schema with two Dimension Connectors with level attribute Level attribute in DimensionConnector uses for optimize sql inner join Level attribute is name of the level to join to If not specified joins to the lowest level of the dimension
Database Schema
The Database Schema contains the Fact table with two columns: DIM_KEY and VALUE. The DATE_KEY column is used as the discriminator in the Hierarchy definitions.
H1_L1
table with two columns:KEY
andNAME
HX_L2
table with 3 columns:KEY
,NAME
,H1L1_KEY
,H2L1_KEY
<roma:DatabaseSchema id="_databaseSchema_optimisationwithlevelattribute">
<tables xsi:type="roma:PhysicalTable" id="_table_fact" name="Fact">
<columns xsi:type="roma:PhysicalColumn" id="_column_fact_dim_key" name="DIM_KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_column_fact_value" name="VALUE" type="Integer"/>
</tables>
<tables xsi:type="roma:PhysicalTable" id="_h1_l1" name="H1_L1">
<columns xsi:type="roma:PhysicalColumn" id="_h1_l1_key" name="KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_h1_l1_name" name="NAME"/>
</tables>
<tables xsi:type="roma:PhysicalTable" id="_hx_l2" name="HX_L2">
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_key" name="KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_name" name="NAME"/>
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_h1l1_key" name="H1L1_KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_h2l1_key" name="H2L1_KEY" type="Integer"/>
</tables>
</roma:DatabaseSchema>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
H1L1Query
The Query is a simple TableQuery that selects all columns from the H1_L1 table to use in the hierarchy join.
<roma:TableQuery id="_h1l1query" table="_h1_l1"/>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
HxL2Query
The Query is a simple TableQuery that selects all columns from the Hx_L2 table to use in the hierarchy join.
<roma:TableQuery id="_hxl2query" table="_hx_l2"/>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
Query
The Query is a simple TableQuery that selects all columns from the Fact table to use in the measures.
<roma:TableQuery id="_table_factQuery" table="_table_fact"/>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
JoinQuery
The JoinQuery specifies which TableQueries should be joined. It also defines the columns in each table that are used for the join:
- In the lower-level table (
Hx_L2
), the join uses the foreign keyH1L1_KEY
. - In the upper-level table (
H1_L1
), the join uses the primary keyKEY
.
<roma:JoinQuery id="_joinQuery">
<left key="_hx_l2_h1l1_key" query="_hxl2query"/>
<right key="_h1_l1_key" query="_h1l1query"/>
</roma:JoinQuery>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
H1_Level1
This Example uses H1_Level1
level bases on the KEY
column and name column NAME
of table H1_L1
.
<roma:Level id="_h1_level1" name="H1_Level1" column="_h1_l1_key" nameColumn="_h1_l1_name"/>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
H1_Level2
This Example uses H1_Level2
level bases on the KEY
column and name column NAME
of table HX_L2
.
<roma:Level id="_h1_level2" name="H1_Level2" column="_hx_l2_key" nameColumn="_hx_l2_name"/>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
Hierarchy1
The Hierarchy1 is defined with the hasAll property set to false and the two levels H1_Level1
and H1_Level2
.
<roma:ExplicitHierarchy id="_hierarchy1" name="Hierarchy1" primaryKey="_hx_l2_key" query="_joinQuery" levels="_h1_level1 _h1_level2"/>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
Diml1
The time dimension is defined with the one hierarchy.
<roma:StandardDimension id="_diml1" name="Diml1" hierarchies="roma:ExplicitHierarchy _hierarchy1"/>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
Cube
The cube with two DimensionConnectors. One of them with level attribute to level H1_Level1 Second of them with level attribute to level H1_Level2 Level attribute in DimensionConnector uses for optimize sql inner join Level attribute is name of the level to join to If not specified joins to the lowest level of the dimension
<roma:PhysicalCube id="_cube" name="Cube" query="_table_factQuery">
<dimensionConnectors foreignKey="roma:PhysicalColumn _column_fact_dim_key" dimension="roma:StandardDimension _diml1" overrideDimensionName="Dim1" level="_h1_level2" id="_dc_dim1"/>
<dimensionConnectors foreignKey="roma:PhysicalColumn _column_fact_dim_key" dimension="roma:StandardDimension _diml1" overrideDimensionName="Dim2" level="_h1_level1" id="_dc_dim2"/>
<measureGroups>
<measures xsi:type="roma:SumMeasure" id="_measure_Measure" name="Measure" column="_column_fact_value"/>
</measureGroups>
</roma:PhysicalCube>
Note: This is only a symbolic example. For the exact definition, see the Definition section.
Definition
This files represent the complete definition of the catalog.
<?xml version="1.0" encoding="UTF-8"?>
<xmi:XMI xmi:version="2.0" xmlns:xmi="http://www.omg.org/XMI" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:roma="https://www.daanse.org/spec/org.eclipse.daanse.rolap.mapping">
<roma:Catalog description="Dimension optimization with level attributes" name="Daanse Tutorial - Dimension Optimisation With Level Attribute" cubes="_cube" dbschemas="_databaseSchema_optimisationwithlevelattribute"/>
<roma:DatabaseSchema id="_databaseSchema_optimisationwithlevelattribute">
<tables xsi:type="roma:PhysicalTable" id="_table_fact" name="Fact">
<columns xsi:type="roma:PhysicalColumn" id="_column_fact_dim_key" name="DIM_KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_column_fact_value" name="VALUE" type="Integer"/>
</tables>
<tables xsi:type="roma:PhysicalTable" id="_h1_l1" name="H1_L1">
<columns xsi:type="roma:PhysicalColumn" id="_h1_l1_key" name="KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_h1_l1_name" name="NAME"/>
</tables>
<tables xsi:type="roma:PhysicalTable" id="_hx_l2" name="HX_L2">
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_key" name="KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_name" name="NAME"/>
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_h1l1_key" name="H1L1_KEY" type="Integer"/>
<columns xsi:type="roma:PhysicalColumn" id="_hx_l2_h2l1_key" name="H2L1_KEY" type="Integer"/>
</tables>
</roma:DatabaseSchema>
<roma:TableQuery id="_h1l1query" table="_h1_l1"/>
<roma:TableQuery id="_hxl2query" table="_hx_l2"/>
<roma:TableQuery id="_table_factQuery" table="_table_fact"/>
<roma:JoinQuery id="_joinQuery">
<left key="_hx_l2_h1l1_key" query="_hxl2query"/>
<right key="_h1_l1_key" query="_h1l1query"/>
</roma:JoinQuery>
<roma:Level id="_h1_level1" name="H1_Level1" column="_h1_l1_key" nameColumn="_h1_l1_name"/>
<roma:Level id="_h1_level2" name="H1_Level2" column="_hx_l2_key" nameColumn="_hx_l2_name"/>
<roma:ExplicitHierarchy id="_hierarchy1" name="Hierarchy1" primaryKey="_hx_l2_key" query="_joinQuery" levels="_h1_level1 _h1_level2"/>
<roma:StandardDimension id="_diml1" name="Diml1" hierarchies="_hierarchy1"/>
<roma:PhysicalCube id="_cube" name="Cube" query="_table_factQuery">
<dimensionConnectors foreignKey="_column_fact_dim_key" dimension="_diml1" overrideDimensionName="Dim1" level="_h1_level2" id="_dc_dim1"/>
<dimensionConnectors foreignKey="_column_fact_dim_key" dimension="_diml1" overrideDimensionName="Dim2" level="_h1_level1" id="_dc_dim2"/>
<measureGroups>
<measures xsi:type="roma:SumMeasure" id="_measure_Measure" name="Measure" column="_column_fact_value"/>
</measureGroups>
</roma:PhysicalCube>
</xmi:XMI>
Tutorial Zip
This files contaisn the data-tables as csv and the mapping as xmi file.