Derived data enables you to reference report results in other reports, in essence "stacking" results from sub-select queries. The derived data feature also supports cohort analysis, where a set of data from a report is used (joined back) in another report, and allows you to build computed columns for re-use.
You must first define derived data, perhaps define additional derived data, and then use it. There also options for saving, viewing and deleting existing derived data measures.
Derived Data is very useful in determining weighted averages and other, more complex
calculations. For example, in the dataset World Life Expectancy
, life
expectancy is reported at the level of each country, for each year. If we wanted to determine
the life expectancy by region or subregion, we have to calculate a weighted average of life
expectancies. You can also parametrize derived data definitions using bracket notation.
The following steps demonstrate how to use derived data on a table visual based on the dataset
World Life Expectancy
[data source
samples.world_life_expectancy
]. The initial set-up follows:
un_region
, un_subregion
and
country
on the Dimension shelf.population
, as sum(population)
, on the
Measures shelf.year
on the Filters shelf, and
change the expression to [year]=<<year_param:2000>>
. This enables you
to dynamically change derived data calculations. You must specify a default value in the
parametrized expression.