![]() Previous |
![]() Next |
Within an aggregation specification, a DROP DIMENSION statement specifies how non-hierarchical aggregation across variables is performed. You use this statement in aggregation specification when you are aggregating the detail data from one variable (the source variable) into another variable (the target variable) and you want to aggregate across a non-hierarchical dimension of the source variable. In this case, the target variable has one less dimension (the "dropped" dimension) than the source variable because the values of the source variable associated with this dimension are aggregated to populate the target variable.
Syntax
DROP DIMENSION dimname [VALUES {valsetname|ALL} OPERATOR operation [ARGS argument]
where argument is one or more of the following phrases:
DIVIDEBYZERO {YES|NO}
DECIMALOVERFLOW {YES|NO}
NASKIP {YES|NO}
WEIGHTBY [WNAFILL {number|NA}] wobj
Parameters
The name of a dimension in the source variable that contains the detail data.
Sets the status of dimname during the aggregation.
The name of a valueset object that determines the status of the dimension specified by dimname.
Specifies that all of the values of dimname are in status.
Identifies the calculation method used to aggregate the data.
A keyword that describes the type of aggregation to perform. The keywords are listed in Table: Aggregation Methods.
Indicates optional handling of the aggregation.
Specifies whether to allow division by zero. Specify YES to allow division by zero which means that a statement involving division by zero executes without error but produces NA
results. Specify NO to disallow division by zero which means that a statement involving division by zero stops executing and produces an error message. The default value is the current value of the DIVIDEBYZERO option.
Specifies whether to allow decimal overflow, which occurs when the result of a calculation is very large and can no longer be represented by the exponent portion of the numeric representation. Specify YES to allow overflow, which means that a calculation that generates overflow executes without error and produces NA
results. Specify NO to disallow overflow which means that a calculation involving overflow stops executing and generates an error message. The default value is the current value of the DECIMALOVERFLOW option.
Specifies whether NA
values are input. Specify YES when you want Oracle OLAP to ignore NA
values when aggregating which means that only actual values are used in calculations. Specify NO when you want Oracle OLAP to consider NA
values when aggregating which means that when any of the values being considered are NA
, the calculation returns NA
. The default value is the current value of the NASKIP option.
The value that you specify for the NASKIP phrase does not effect calculation performed when you specify HAVERAGE, HFIRST, HLAST, HWAVERAGE, HWFIRST, HWLAST for operation.
Indicates that weighted aggregation is to be performed. You must include a WEIGHTBY clause when you specify HWAVERAGE, HWFIRST, HWLAST, SSUM, WAVERAGE, WFIRST, WLAST, or WSUM for operation. The WEIGHTBY phrase always includes a wobj argument and can optionally include the WNAFILL keyword. For more information about the use of the WEIGHTBY phrase, see the RELATION (for aggregation) statement of the AGGMAP command.
Indicates handling for NA
values. The default values for WNAFILL vary depending on the value of operation. For more information about using the WNAFILL phrase, see the RELATION (for aggregation) statement of the AGGMAP command.
Substitutes a number for every NA
value. That number replaces every NA
value in the weight object, weight formula, or weight relation.
0.0
is the default for HWAVERAGE and SSUM.
1.0
is the default for HWFIRST, HWLAST, WAVERAGE, WFIRST, WLAST, and WSUM.
Specifies that NA
values are to be specified as NA
. NA
is the default for OR.
A variable, formula, or relation that provides the weighted values. It can be numeric or BOOLEAN. When wobj is BOOLEAN, then TRUE
has a weight of 1.0
and FALSE
has a weight of 0.0
. A formula is queried only when needed, depending on the dimensionality of the formula and the variable being aggregated. When wobj is a relation, it should be a one-dimensional self-relation. For more information about specifying values for wobj, see the RELATION (for aggregation) statement of the AGGMAP command.
Examples
For an example of using a DROP DIMENSION statement in an aggregation specification, see Example: Aggregating into a Different Variable.