A large volume and variety of data generally need data profiling to understand the nature of data. One of the aspects of data is hierarchy and inter-relationships within different attributes in data. Hierarchical data is often nested at multiple levels. To analyze the relationship between different attributes in a data that is hierarchical, drill-down and drill-through are two of the most common techniques that are employed for data exploration as well as use-cases like root cause analysis. While these techniques are standard and have been in the industry for quite a long time, figuring out these relationships and navigating hierarchical data can be a challenging task. Data Analysts or Business Analysts typically perform this analysis on the data before presenting it to the end-users. In certain cases, some domain or business users may be required to perform such analysis on the report itself. In that case, the task becomes even more challenging considering the limited data analysis capabilities offered by a reporting tool compared to a database and query languages like SQL. To help power users perform such analysis on a reporting tool, visualizations like decomposition trees can be used to decompose hierarchical data that is presented in an aggregated manner. The Decomposition tree can support both drill-down as well as drill-through use-cases when the user is provided the flexibility to choose the hierarchy or dimensions on-demand. In the Microsoft technology stack, Power BI is the key reporting tool for authoring reports and supports a wide variety of data sources. Power BI offers a category of visuals which are known as AI visuals. One such visual in this category is the Decomposition Tree.
Read on to see how you can create a decomposition tree, what kind of information it shows, and how you can interact with it to learn more about correlations and causes.