Applies to:  CELONIS 4.0 CELONIS 4.2 CELONIS 4.3 CELONIS 4.4 CELONIS 4.5 CELONIS 4.6 CELONIS 4.7 

Description

In general, only one aggregation is calculated for a query. With a global aggregation it is possible to calculate an additional aggregation function over all values of a column.

GLOBAL calculates an aggregation function for an aggregation to which all values are aggregated into one group.

Like normal aggregations, a global aggregation function take filters and selections into account. Values which are filtered out are not part of the result. Therefore if a filter or a selection is changed, every global aggregation function is recalculated.

Syntax

GLOBAL ( aggregation )

NULL handling

If the input value is NULL, then GLOBAL returns NULL as well.

Note

The result of GLOBAL can not be used as a grouper column for an aggregation.

Like the standard aggregation functions, GLOBAL can not be used inside a FILTER statement.

Examples


[1] Calculating the sum for two groups and the count for one group

Query
Column1
"Table1"."Dimension"
Column2
SUM ( "Table1"."Value" ) / GLOBAL ( COUNT_TABLE ( "Table1" ) )
Input
Table1
Dimension : STRINGValue : FLOAT
'A'
1.0
'B'
1.0
'A'
1.0
'B'
1.0
Output
Result
Column1 : STRINGColumn2 : FLOAT
'A'
0.5
'B'
0.5



[2] Calculating the sum for two groups and the count for one group as a result of a filter

Query
Filter
FILTER "Table1"."Dimension" != 'C';
Column1
"Table1"."Dimension"
Column2
SUM ( "Table1"."Value" ) / GLOBAL ( COUNT ( "Table1"."Value" ) )
Input
Table1
Dimension : STRINGValue : FLOAT
'A'
1.0
'B'
1.0
'C'
1.0
'A'
1.0
'B'
1.0
'C'
1.0
Output
Result
Column1 : STRINGColumn2 : FLOAT
'A'
0.5
'B'
0.5



[3] For each row, return 1 if the value is greater than the global average value (5) of that column:

Query
Column1
"Table1"."Dimension"
Column2
CASE WHEN "Table1"."Value" > GLOBAL ( AVG ( "Table1"."Value" ) ) THEN 1 ELSE 0 END
Input
Table1
Dimension : STRINGValue : INT
'A'
2
'B'
4
'C'
6
'D'
8
Output
Result
Column1 : STRINGColumn2 : INT
'A'
0
'B'
0
'C'
1
'D'
1


See also:

  • No labels