Level of Detail Part 1: Introduction

by Peter Silvester

Before embarking on any kind of data analysis or visualisation it important that you know your data and the structure contained within it. This is particularly key when we are working with Level of Detail calculations because in order to specify the level at which calculations are done we need to know how the hierarchies of our data.

The dimensions within our data determine the level of granularity or detail that describes the data. This is easiest to demonstrate using geographical hierarchies. If our data set only contained these dimensions then the lowest level of detail would be Post Code and the highest (below total aggregation) is Country.

 

granularity

Level of Detail in Tableau

When working in Tableau we change the Level of Detail in our view by dragging in Dimensions from the data pane. The diagram below shows where you need to drop your dimensions for them to have an effect on the view LoD. Dimensions on areas shaded green such as the Columns or Rows or the majority of the marks card will affect the LoD, whereas Filters or Tooltips do not.

 

changing-the-view

 

Placing measures into the view will create aggregations based on the dimensions you have selected. However there will be times when you want to calculate aggregations based on detail you determine and not one the view level of detail. This is where Tableau’s Level of Detail Calculations can be used to change to your desired level of aggregation. Calculations can be written as ‘FIXED’ at a particular level of aggregation or you can choose to ‘INCLUDE’ or ‘EXCLUDE’ dimensions. For a full explanation of the how to use the different keywords and the syntax required for LoD Calculations see part 2 of the LoD Series (coming soon).

Avatar

Peter Silvester

Thu 01 Dec 2016

Tue 22 Nov 2016

© 2022 The Information Lab Ltd. All rights reserved.