Decomposition tree visual – Performing Advanced Analysis

Decomposition tree visual

Similar to the key influencers visualization, the decomposition tree visual also lets you see your data across multiple dimensions. You can use it for improvised exploration and conducting root cause analysis.

The decomposition tree visualization can be found in the Visualizations pane. Once you have added it to the report, you start by selecting the field you want to analyze. You can then add one, or preferably more, fields to Explain by. If this sounds familiar, it’s because it’s the exact same thing you did for the key influencers visual.

Figure 13.13 – You can choose the order you want to decompose your tree in

The decomposition tree visual will update as data fields are added to the Explain by configuration. When a new field is added to Explain by, the visual updates, with the + symbol being added to the visual. By clicking +, you can then select how you want the decomposition tree visual to generate the next level of understanding from the data. When the AI computation in the visual detects high values and low values in the data, this can also be selected and is referred to as an AI split. These are provided by the visual to help you better analyze the data using the decomposition tree visual.

Applying AI insights

When you have a dataset and don’t know where to start, there is a feature that can help you. Quick insights is a machine learning-based capability built into Power BI that can automatically build visualizations and make some insights stand out. It is also useful to use on datasets that have been published but allows you to look for any insights that may have been missed. This capability can be used on some datasets that have been published to a Power BI workspace.

To use quick insights, click Get quick insights from the dataset menu for a dataset that has been published to the Power BI service. The service will then run statistical algorithms to automatically generate visualizations.

Figure 13.14 – After clicking on Get quick insights, a popup will appear telling you that your insights are ready

Once the process has completed, you need to click View insights to view the insights that have been generated by the service. When quick insights are generated, Power BI will create up to 32 different insight cards. Each card will have a visual and a short description. These visuals are the same as visuals you manually create. These visuals can be pinned to dashboards the same as visuals in reports.

Figure 13.15 – The output from quick insights. Power BI generated 37 visuals this time

Quick insights can give you a deeper understanding of your data and the potentially hidden relationships within. It’s like having your dataset reviewed by a data scientist, but at a fraction of the cost!