To make a cube functional, you need to identify the relevant dimensions and measures and create them in the cube.
To create a Cube, apply the following steps:
This example shows you how to quickly create a simple cube in a report to export its measures.
The implementation steps are detailed below. Exhaustive options and descriptions are available in the other sections of this chapter.
To create a cube, click the New button above the list of cubes.
Select the fact schema, i.e. the schema which contains the elements you want to explore. In this example, we are going to select the Recipient table.
Click Save to create the Cube: it will appear on the list of Cubes and may then be configured using the appropriate tabs.
Click the Filter the source data… link to apply the calculations of this Cube to a select of data in the database.
Dimensions coincide with analysis axes defined for each Cube based on their related fact schema. These are the dimensions explored in the analysis, such as time (year, month, date…), a classification of products or contracts (family, reference, etc.), a population segment (by city, age group, status, etc.).
These analysis axes are defined in the Dimension tab of the Cube.
Click the Add button to create a new dimension, then in the Expression field, click the Edit expression icon to select the field that contains the concerned data.
Start by selecting the recipient Age. For this field, you can define binning to group ages and make information reading easier. We recommend using binning when there is a likelihood of several separate values.
To do this, check the Enable binning option. Binning modes are detailed in Data binning.
Add a Date type dimension. Here, we want to display recipient profile creation dates
To do this, click Add and select the Creation date field in the recipient table.
It’s possible to select the date display mode. To do this, select the hierarchy to use and the levels to generate:
In our example, we only want to display years, months and days since it’s not possible to work with weeks and semesters/months at the same time: these levels aren’t compatible.
Create another dimension to analyze data relative to the recipient’s city
To do this, add a new dimension and select the city in the Location node of the recipient schema.
You can enable binning to make information reading easier and link the values to an enumeration.
Select the enumeration from the drop-down list
Only the values in the enumeration will be displayed. The others will be grouped under the label defined in the Label of the other values field.
For more on this, refer to Dynamically managing bins.
Once the dimensions are defined, you need to specify a calculation mode for the values to be displayed in the cells. To do this, create the matching indicators in the Measures tab: create as many measures as there are columns to display in the report which will use the cube.
To do this, apply the following steps:
Click the Add button.
Select the type of measure and the formula to apply. Here we want to count the number of women among the recipients.
Our measure is based on the fact schema and uses the Count operator.
The Filter the measure data… link lets you select only women. For more on defining measures and the available options, refer to Defining measures.
Enter the label of the measure and save it.
Save the cube.
Once the cube is configured, it can be used as a template for creating a new report.
To do this:
Click the Create button of the Reports tab and select the cube you have just created.
Click the Create button to confirm: this will take you to the report configuration and viewing page.
By default, the first two available dimensions are offered in lines and columns but no value is displayed in the table. To generate the table, click the main icon:
You can switch the axes of the dimension, delete them, add new measures, etc. Possible operations are detailed here: Using cubes to explore data.
To do this, use the appropriate icons.