[CALCITE-428] Use optimization algorithm to suggest which tiles of a lattice to materialize

• Type: New Feature
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.9.1-incubating
• Assignee: Julian Hyde
• Reporter: Julian Hyde
• Created: 30/Sep/14 20:34
• Updated: 18/Oct/14 22:36
• Resolved: 30/Sep/14 22:33

Description

Use an optimization algorithm to suggest which tiles of a lattice to materialize.

The standard algorithm is described in Implementing data cubes efficiently; Harinarayan, Rajaraman, Ullman 1996, and pentaho-aggdesigner implements it.

Add 'algorithm' and 'rowCountEstimate' attributes to the lattice JSON element:

{
  name: 'star',
  algorithm: true,
  rowCountEstimate: 86000,
  sql: "select 1 from \"foodmart\".\"sales_fact_1997\" as \"s\"\njoin \"foodmart\".\"product\" as \"p\" using (\"product_id\")\njoin \"foodmart\".\"time_by_day\" as \"t\" using (\"time_id\")\njoin \"foodmart\".\"product_class\" as \"pc\" on \"p\".\"product_class_id\" = \"pc\".\"product_class_id\"",
  auto: false,
  defaultMeasures: [ {
      agg: 'sum',
      args: 'unit_sales'
    }, {
      agg: 'sum',
      args: 'store_sales'
    }, {
      agg: 'count'
  } ],
  tiles: [ {
    dimensions: [ 'the_year', ['t', 'quarter'] ],
    measures: [ ]
  } ]
}

algorithm enables the optimization algorithm.

rowCountEstimate specifies an estimate of the number of rows in the lattice. If not specified, 1000 is used.