var
The var
function calculates the sample variance of the set of
numbers in the specified measure, grouped by the chosen dimension or
dimensions.
Syntax
var(measure, [group-by level])
Arguments
- measure
-
The argument must be a measure. Null values are omitted from the results. Literal values don't work. The argument must be a field.
- group-by level
-
(Optional) Specifies the level to group the aggregation by. The level added can be any dimension or dimensions independent of the dimensions added to the visual.
The argument must be a dimension field. The group-by level must be enclosed in square brackets
[ ]
. For more information, see LAC-A functions.
Examples
The following example returns the variance of a sample of test scores.
var({Scores})
You can also specify at what level to group the computation using one or more dimensions in the view or in your dataset. This is called a LAC-A function. For more information about LAC-A functions, see LAC-A functions. The following example returns the variance of a sample of test scores at the subject level, but not across other dimensions (Class) in the visual.
var({Scores}, [Subject]