OpenText Analytics Database 26.2.x – GROUP BY clause

Sql-Reference: CUBE aggregate

Mon, 01 Jan 0001 00:00:00 +0000

Automatically performs all possible aggregations of the specified columns, as an extension to the GROUP BY clause.

You can use the ROLLUP clause with three grouping functions:

Syntax

GROUP BY group-expression[,...]

Parameters

group-expression``

One or both of the following:

An expression that is not an aggregate or a grouping function that includes constants and column references in FROM-specified tables. For example:
```
column1, (column2+1), column3+column4
```
A multilevel expression, one of the following:
- ROLLUP
- CUBE
- GROUPING SETS

Restrictions

GROUP BY CUBE does not order data. If you want to sort data, use the ORDER BY clause. The ORDER BY clause must come after the GROUP BY clause.
You can use CUBE inside a GROUPING SETS expression, but not inside a ROLLUP expression or another CUBE expression.

Levels of CUBE aggregation

If n is the number of grouping columns, CUBE creates 2ⁿ levels of aggregations. For example:

CUBE (A, B, C) creates all possible groupings, resulting in eight groups:

(A, B, C)
(A, B)
(A, C)
(B, C)
(A)
(B)
(C)
()

If you increase the number of CUBE columns, the number of CUBE groupings increases exponentially. The CUBE query may be resource intensive and produce combinations that are not of interest. In that case, consider using theGROUPING SETS aggregate, which allows you to choose specific groupings.

Examples

Using CUBE to return all groupings

Suppose you have a table that contains information about family expenses for books and electricity:

=> SELECT * FROM expenses ORDER BY Category, Year;
 Year |  Category   | Amount
------+-------------+--------
2005  | Books       |  39.98
2007  | Books       |  29.99
2008  | Books       |  29.99
2005  | Electricity | 109.99
2006  | Electricity | 109.99
2007  | Electricity | 229.98

To aggregate the data by both Category and Year using the CUBE aggregate:

=> SELECT Category, Year, SUM(Amount) FROM expenses
   GROUP BY CUBE(Category, Year) ORDER BY 1, 2, GROUPING_ID();
 Category    | Year |  SUM
-------------+------+--------
 Books       | 2005 |  39.98
 Books       | 2007 |  29.99
 Books       | 2008 |  29.99
 Books       |      |  99.96
 Electricity | 2005 | 109.99
 Electricity | 2006 | 109.99
 Electricity | 2007 | 229.98
 Electricity |      | 449.96
             | 2005 | 149.97
             | 2006 | 109.99
             | 2007 | 259.97
             | 2008 |  29.99
             |      | 549.92

The results include subtotals for each category and year, and a grand total ($549.92).

Using CUBE with the HAVING clause

This example shows how you can restrict the GROUP BY results, use the HAVING clause with the CUBE aggregate. This query returns only the category totals and the full total:

=> SELECT Category, Year, SUM(Amount) FROM expenses
   GROUP BY CUBE(Category,Year) HAVING GROUPING(Year)=1;
 Category    | Year |  SUM
-------------+------+--------
 Books       |      |  99.96
 Electricity |      | 449.96
             |      | 549.92

The next query returns only the aggregations for the two categories for each year. The GROUPING ID function specifies to omit the grand total ($549.92):

=> SELECT Category, Year, SUM (Amount) FROM expenses
   GROUP BY CUBE(Category,Year) HAVING GROUPING_ID(Category,Year)<2
   ORDER BY 1, 2, GROUPING_ID();
 Category   | Year |  SUM
------------+------+--------
 Books      | 2005 |  39.98
 Books      | 2007 |  29.99
 Books      | 2008 |  29.99
 Books      |      |  99.96
 Electrical | 2005 | 109.99
 Electrical | 2006 | 109.99
 Electrical | 2007 | 229.98
 Electrical |      | 449.96

Sql-Reference: GROUPING SETS aggregate

Mon, 01 Jan 0001 00:00:00 +0000

The GROUPING SETS aggregate is an extension to the GROUP BY clause that automatically performs subtotal aggregations on groupings that you specify.

You can use the GROUPING SETS clause with three grouping functions:

To sort data, use the ORDER BY clause. The ORDER BY clause must follow the GROUP BY clause.

Syntax

GROUP BY group-expression[,...]

Parameters

group-expression``

One or both of the following:

An expression that is not an aggregate or a grouping function that includes constants and column references in FROM-specified tables. For example:

column1, (column2+1), column3+column4
A multilevel expression, one of the following:
- ROLLUP
- CUBE
- GROUPING SETS

Defining the groupings

GROUPING SETS allows you to specify exactly which groupings you want in the results. You can also concatenate the groupings as follows:

The following example clauses result in the groupings shown.

This clause...	Defines groupings...
`...GROUP BY GROUPING SETS(A,B,C,D)...`	(A), (B), (C), (D)
`...GROUP BY GROUPING SETS((A),(B),(C),(D))...`	(A), (B), (C), (D)
`...GROUP BY GROUPING SETS((A,B,C,D))...`	(A, B, C, D)
`...GROUP BY GROUPING SETS(A,B),GROUPING SETS(C,D)...`	(A, C), (B, C), (A, D), (B, C)
`...GROUP BY GROUPING SETS((A,B)),GROUPING SETS(C,D)...`	(A, B, C), (A, B, D)
`...GROUP BY GROUPING SETS(A,B),GROUPING SETS(ROLLUP(C,D))...`	(A,B), (A,B,C), (A,B,C,D)
`...GROUP BY A,B,C,GROUPING SETS(ROLLUP(C, D))...`	(A, B, C, D), (A, B, C), (A, B, C) The clause contains two groups (A, B, C). In the HAVING clause, use the GROUP_ID function as a predicate, to eliminate the second grouping.

Example: selecting groupings

This example shows how to select only those groupings you want. Suppose you want to aggregate on columns only, and you do not need the grand total. The first query omits the total. In the second query, you add () to the GROUPING SETS list to get the total. Use the ORDER BY clause to sort the results by grouping:

=> SELECT Category, Year, SUM(Amount) FROM expenses
   GROUP BY GROUPING SETS((Category, Year), (Year))
   ORDER BY 1, 2, GROUPING_ID();
  Category  | Year |  SUM
------------+------+--------
 Books      | 2005 |  39.98
 Books      | 2007 |  29.99
 Books      | 2008 |  29.99
 Electrical | 2005 | 109.99
 Electrical | 2006 | 109.99
 Electrical | 2007 | 229.98
            | 2005 | 149.97
            | 2006 | 109.99
            | 2007 | 259.97
            | 2008 |  29.99
=> SELECT Category, Year, SUM(Amount) FROM expenses
   GROUP BY GROUPING SETS((Category, Year), (Year), ())
   ORDER BY 1, 2, GROUPING_ID();
  Category  | Year |  SUM
------------+------+--------
 Books      | 2005 |  39.98
 Books      | 2007 |  29.99
 Books      | 2008 |  29.99
 Electrical | 2005 | 109.99
 Electrical | 2006 | 109.99
 Electrical | 2007 | 229.98
            | 2005 | 149.97
            | 2006 | 109.99
            | 2007 | 259.97
            | 2008 |  29.99
            |      | 549.92

Sql-Reference: ROLLUP aggregate

Mon, 01 Jan 0001 00:00:00 +0000

Automatically performs subtotal aggregations as an extension to the GROUP BY clause. ROLLUP performs these aggregations across multiple dimensions, at different levels, within a single SQL query.

You can use the ROLLUP clause with three grouping functions:

Syntax

ROLLUP grouping-expression[,...]

Parameters

group-expression

One or both of the following:

An expression that is not an aggregate or a grouping function that includes constants and column references in FROM-specified tables. For example:

column1, (column2+1), column3+column4
A multilevel expression, one of the following:
- ROLLUP
- CUBE
- GROUPING SETS

Restrictions

GROUP BY ROLLUP does not sort results. To sort data, an ORDER BY clause must follow the GROUP BY clause.

Levels of aggregation

If n is the number of grouping columns, ROLLUP creates n+1 levels of subtotals and grand total. Because ROLLUP removes the right-most column at each step, specify column order carefully.

Suppose that ROLLUP(A, B, C) creates four groups:

(A, B, C)
(A, B)
(A)
()

Because ROLLUP removes the right-most column at each step, there are no groups for (A, C) and (B, C).

If you enclose two or more columns in parentheses, GROUP BY treats them as a single entity. For example:

ROLLUP(A, B, C) creates four groups:
```
(A, B, C)
(A, B)
(A)
()
```
ROLLUP((A, B), C) treats (A, B) as a single entity and creates three groups:
```
(A, B, C)
(A, B)
()
```

Example: aggregating the full data set

The following example shows how to use the GROUP BY clause to determine family expenses for electricity and books over several years. The SUM aggregate function computes the total amount of money spent in each category per year.

Suppose you have a table that contains information about family expenses for books and electricity:

=> SELECT * FROM expenses ORDER BY Category, Year;
 Year |  Category   | Amount
------+-------------+--------
2005  | Books       |  39.98
2007  | Books       |  29.99
2008  | Books       |  29.99
2005  | Electricity | 109.99
2006  | Electricity | 109.99
2007  | Electricity | 229.98

For the expenses table, ROLLUP computes the subtotals in each category between 2005–2007:

Books: $99.96
Electricity: $449.96
Grand total: $549.92.

Use the ORDER BY clause to sort the results:

=> SELECT Category, Year, SUM(Amount) FROM expenses
   GROUP BY ROLLUP(Category, Year) ORDER BY 1,2, GROUPING_ID();
 Category    | Year |  SUM
-------------+------+--------
 Books       | 2005 |  39.98
 Books       | 2007 |  29.99
 Books       | 2008 |  29.99
 Books       |      |  99.96
 Electricity | 2005 | 109.99
 Electricity | 2006 | 109.99
 Electricity | 2007 | 229.98
 Electricity |      | 449.96
             |      | 549.92

Example: using ROLLUP with the HAVING clause

This example shows how to use the HAVING clause with ROLLUP to restrict the GROUP BY results. The following query produces only those ROLLUP categories where year is subtotaled, based on the expression in the GROUPING function:

=> SELECT Category, Year, SUM(Amount) FROM expenses
   GROUP BY ROLLUP(Category,Year) HAVING GROUPING(Year)=1
   ORDER BY 1, 2, GROUPING_ID();
 Category    | Year |  SUM
-------------+------+--------
 Books       |      |  99.96
 Electricity |      | 449.96
             |      | 549.92

The next example rolls up on (Category, Year), but not on the full results. The GROUPING_ID function specifies to aggregate less than three levels:

=> SELECT Category, Year, SUM(Amount) FROM expenses
   GROUP BY ROLLUP(Category,Year) HAVING GROUPING_ID(Category,Year)<3
   ORDER BY 1, 2, GROUPING_ID();
 Category    | Year |  SUM
-------------+------+--------
 Books       | 2005 |  39.98
 Books       | 2007 |  29.99
 Books       | 2008 |  29.99
 Books       |      |  99.96
 Electricity | 2005 | 109.99
 Electricity | 2006 | 109.99
 Electricity | 2007 | 229.98
 Electricity |      | 449.96

OpenText Analytics Database 26.2.x – GROUP BY clause

Sql-Reference: CUBE aggregate

Syntax

Parameters

Restrictions

Levels of CUBE aggregation

Examples

Using CUBE to return all groupings

Using CUBE with the HAVING clause

See also

Sql-Reference: GROUPING SETS aggregate

Syntax

Parameters

Defining the groupings

Example: selecting groupings

See also

Sql-Reference: ROLLUP aggregate

Syntax

Parameters

Restrictions

Levels of aggregation

Example: aggregating the full data set

Example: using ROLLUP with the HAVING clause

See also