Explain the increase/decrease

Quick Insight features are enabled in Power BI Desktop by default, allowing users to right-click data points in visuals and execute the relevant analysis. In the following example, the user has right-clicked the data point for 2017-Apr, and as a result, an option to explain the decrease is exposed in the Analyze menu: 

Explaining the decrease in Power BI Desktop

Clicking Explain the decrease executes machine learning algorithms against the dataset and populates a window with visuals representing the insights retrieved. The user can scroll vertically to view the different insights obtained such as the Customer Gender column accounting for a majority of the decrease, or Product Name XYZ, which had the largest decrease among all products.

By default, a waterfall visual is used to display each insight, but other visuals such as the scatter chart and the 100 percent stacked column chart are available too. In the following example, the user has scrolled to an insight based on the Customer History Segment column and views the data as a waterfall chart:

Explain the decrease in Power BI Desktop

Clicking the plus sign at the top right corner of the text box explaining the insight adds the visual to the report page. Adding the visual to the report page automatically populates the associated field wells and visual level filters as though the visual was created manually. If necessary, the report author can apply further formatting to align the visual with the design and layout of the page. 

Currently, Quick Insights in Power BI Desktop is limited to the local dataset and is exclusive to import mode datasets. For example, the Explain the decrease option will not appear when connecting to a published Power BI dataset or a SSAS database via Live connection. Given the importance of isolating reports from a central dataset as described in the previous chapter, Chapter 5Creating and Formatting Power BI Reports this limitation represents a significant obstacle to utilize this feature in corporate deployments.

Additionally, there are several limitations on the kinds of measures and filters supported. For example, measures which use the DISTINCTCOUNT() and SUMX() functions are not supported, and measures containing conditional logic (for example, IF()) cannot be either.
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
3.137.164.24