Tutorial/Schema cheatsheet

From TempusServa wiki
Jump to navigation Jump to search

Introduction

The below paragraphs contains sample snippets for your OLAP schema.

Precise documentation on how to define a schema, can be found at: http://mondrian.pentaho.com/documentation/schema.php

Schema structure overview

  • Table (one):
    • Structure: The centre table containing the measures (FACT table)
    • OLAP cube: Not displayed
  • Dimensions (many)
    • Structure: Related tables or groupable values (FACT table relations)
    • OLAP cube: Columns/row "headers" in the cube
  • Measures (many)
    • Structure: Values found in the centre table (FACT table values)
    • OLAP cube: Numbers to be displayed in the table cells

Defining dimensions for related values

Standard lookup value

The cube schema below only needs adjustment for the system name of the lookup field.

Structure

Solution as displayed "Sample solution" : "Category" = xxx
Solution system names sample : CATEGORY = xxx
Database table names data_sample : CATEGORY = xxx

Data model

  • data_sample
    • CATEGORY: Field containing the lookup value

Cube schema

   ...
... <Dimension name="Example" foreignKey="CATEGORY"> <Hierarchy hasAll="true" primaryKey="LookupID">
<Level name="Category" column="Value" uniqueMembers="true"/> </Hierarchy> </Dimension> ...

Simple choice value

The cube schema below only needs adjustment for the system name of the lookup field.

Structure

Solution as displayed "Sample solution" : "My choice" = xxx
Solution system names sample : CHOICE = xxx
Database table names data_sample : CHOICE = xxx

Data model

  • data_sample
    • CHOICE: Field containing the lookup value

Cube schema

   ...
... <Dimension name="Example" foreignKey="CHOICE"> <Hierarchy hasAll="true" primaryKey="ChoiceID">
<Level name="Answer" column="Value" uniqueMembers="true"/> </Hierarchy> </Dimension> ...

Standard record Status

The cube schema below can be copied directly without modification: The status properties and tablenames are allways the same.

Structure

Solution as displayed "Sample solution" : "Status" = xxx
Solution system names sample : "StatusID" = xxx
Database table names data_sample "StatusID" = xxx

Data model

  • data_sample
    • StatusID: Field containing status reference (allways the same)

Cube schema

   ...
... <Dimension name="Example" foreignKey="StatusID"> <Hierarchy hasAll="true" primaryKey="StatusID">
<Level name="Status" column="Status" uniqueMembers="true"/> </Hierarchy> </Dimension> ...

Defining dimensions for related records

Related solution ONE step away

Structure

Solutions as displayed "Some child" : "Parent" -> "Father or mother"
Solution system names child : PARENT -> parent
Database table names data_child : PARENT -> data_parent

Data model

  • data_child
    • PARENT: Key to the "parent" solution
  • data_parent
    • GRANDPARENT: Key to the "grandparent" solution
    • PARENTNAME: Descriptive field

Cube schema

   ...
... <Dimension name="Example" foreignKey="PARENT"> <Hierarchy hasAll="true" primaryKey="DataID" primaryKeyTable="data_parent">
<Level name="Parent" column="PARENTNAME" uniqueMembers="true"/> </Hierarchy> </Dimension> ...

Related solution TWO steps away

Note that the schemas for multi join tables are written from "inside out", that might seem counterintuitive i relation to what you want to display in the cube later.

Structure

Solutions as displayed "Some child" : "Parent" -> "Father or mother"  : "Grand parent" -> "Grandma and Grandpa's"
Solution system names child : PARENT -> parent : GRANDPARENT -> grandparent
Database table names data_child : PARENT -> data_parent : GRANDPARENT -> data_grandparent

Data model

  • data_child
    • PARENT: Key field pointing to the "parent" solution
  • data_parent
    • GRANDPARENT: Key field pointing to the "grandparent" solution
    • PARENTNAME: Descriptive field
  • data_grandparent
    • GRANDPARENTNAME: Descriptive field

Cube schema

   ...
... <Dimension name="Example" foreignKey="PARENT"> <Hierarchy hasAll="true" primaryKey="DataID" primaryKeyTable="data_parent"> <Join leftKey="GRANDPARENT" rightKey="DataID">
</Join> <Level name="Grandparent" table="data_grandparent" column="GRANDPARENTNAME" uniqueMembers="true"/> <Level name="Parent" table="data_parent" column="PARENTNAME" uniqueMembers="true"/> </Hierarchy> </Dimension> ...

Defining dimensions for inline values

Text values

Cube schema

   ...
   <Dimension name="Category">
       <Hierarchy hasAll="true">
           <Level name="Name" column="NAME" uniqueMembers="true"/>
       </Hierarchy>
   </Dimension>
   ...


Year/integer values

Cube schema

   ...
   <Dimension name="Period" type="TimeDimension">
       <Hierarchy hasAll="true">
           <Level name="Aar" column="YEAR" type="Numeric" uniqueMembers="false" levelType="TimeYears"/>
       </Hierarchy>
   </Dimension>
   ...

Date / datetime values

Cube schema

   ...
   <Dimension name="Period" type="TimeDimension">
     <Hierarchy name="Periode" hasAll="true" allMemberName="All period">
<Level name="Aar" levelType="TimeYears" uniqueMembers="true"> <KeyExpression> <SQL dialect="mysql">Year(CreatedAt)</SQL> <SQL dialect="generic">YEAR</SQL> </KeyExpression> </Level> <Level name="Maaned" uniqueMembers="false" levelType="TimeMonths"> <KeyExpression> <SQL dialect="mysql">Month(CreatedAt)</SQL> <SQL dialect="generic">Month</SQL> </KeyExpression> </Level> </Hierarchy> </Dimension> ...

Enumeration values

Cube schema

   ...
   <Dimension name="Severity">
       <Hierarchy hasAll="true" primaryKey="SEVERITY">
           <InlineTable alias="enumeration">
           <ColumnDefs>
               <ColumnDef name="id" type="Numeric"/>
               <ColumnDef name="desc" type="String"/>
           </ColumnDefs>
           <Rows>
               <Row>
                   <Value column="id">1</Value>
                   <Value column="desc">High</Value>
               </Row>
               <Row>
                   <Value column="id">2</Value>
                   <Value column="desc">Medium</Value>
               </Row>

... more values ...

           </Rows>
       </InlineTable>
       <Level name="Severity" column="id" nameColumn="desc" uniqueMembers="true"/>
   </Hierarchy>
   </Dimension>
   ...

Defining measures

Normal values

Measures are allways numeric values, that can be agggated to higher levels (the levels in the dimensions)

Type Aggregator Examples
Sums SUM Time spent, costs
Average AVG Process time


Cube schema

Calculated values