Understanding dataset attribute, measure and pivot columns

Dashboard components are based on information calculated from a dataset. The creation of components of all types involves selecting columns from the chosen dataset. 
Dataset columns can be of three types: AttributeMeasure and Pivot. Attributes appear within the dataset with blue column headers whereas measures appear with red column headers. Pivot columns are visualised in green.

sc18
 

The above screenshot shows a chart created from an attribute column 'Accounting Period' and a measure column 'Debit Value'. Just drag-and-drop the attribute from the dataset onto the chart followed  by the measure(s).

Attributes and Measures (including Pivots) are defined within the dataset builder.

Attribute columns contain descriptive information about the underlying data - e.g., names, dates, identifiers, descriptions etc.
Measure columns provide results that are calculated from the underlying column values - e.g., sum(Debit), CountDistinct(CustomerID).  In a chart, the attribute provides the x-axis values and the measure the y-axis ones as in the above example.

Pivot columns are special types of measures. Pivots allow you to create more dynamic components - by automatically 'pivoting'Measure by an Attribute (String or DateTime type) column.

Attributes and measures can be used in various combinations in differing component types. One of the main advantages of the Intuitive Dashboards software is that it does not require the dashboard designer to understand the derivations of the attribute and measure columns within the chosen dataset. 

Measure columns can contain sophisticated mathematical expressions such as ( sum(sales) * TotalCountDistinct(CustomerID) ) / ( CountDistinct(CustomerID) * TotalSum(Sales) ).
 
For readers familiar with common Business Intelligence terminology, Attributes would typically be found within dimension tables in a BI schema whereas measures would typically be found in fact tables.



Eg. Within the dataset column expression editor, if you create a measure called 'Gross Total' and  'pivot' the 'Gross Total' by the 'Status' attribute column, you've created a pivot column that can automatically show the 'Gross Total' by 'Status' on a chart. Each individual status will be shown dynamically as an individual series on a chart (or column in a grid).
 
Pivot columns are defined in the dataset column expression editor:

sc12

Now you can then create a component to show dynamically the sum of the 'Gross Total' by 'Status' by 'Supplier Name'. You can also combine Measures and a Pivot column on one component.


Example Pivot Chart: Supplier gross totals by status (Stacked) sc13

Note: You can use only one Pivot column at a time on a grid or a chart component. You cannot use a Pivot on a gauge component.
When building the component, you can choose to display your pivot column in a grid or as a stacked or unstacked bar chart, depending on how you would consider it more readable.

Example Chart: Supplier gross totals by status (Unstacked)

sc14

 

Example Grid: Supplier gross totals by status

sc15

This Grid and charts above only contains three columns from the dataset:

  • Supplier name (Attribute)
  • Gross total (Measure)
  • Gross total by status (Pivot)