When we create any visual, it is very common that we experience increase or decrease in values. We did not know the reason about the cause of these fluctuations. Using insights in Power BI Desktop we can understand the cause of these variation in few clicks by using Analyze option in Power BI.
Let’s create a bar chat to shows Sales by Year.
In this visual we can see decrease in 2015 year and increase in next year.
We can ask Power BI desktop to explain the points those matters for increase and decrease in bar chart. We can get automated insights with full analysis of our data set.
To open Analyze option: Right click on bar of chart > select Analyze > we get two option one is Explain the Decrease and second one is Find where this distribution is different.
Select Explain the Decrease and new insights are open in a new window.
Here we have full analysis of the decrease in Sales and Power BI has Evaluates 14 factors those matter for decrease in Sales. We can see all factors with visualization by scroll down option.
Let’s take 1st one i.e. Sales by Year and Sub Category. It has explained that ‘Machines’ and ‘Supplies’ accounted for the majority of decrease among Sub Category, Offsetting the increase of ‘Bookcases’. The relative contribution made by ‘Machines’, ‘Bookcases’, ‘Accessories’ changed the most.
By selecting the small icons at the bottom of the waterfall visual, we can choose to have insights display a scatter chart, stacked column chart, or a ribbon chart.
Go to Sales by Year and Region visualization and from thumbnail’s change waterfall viz to Ribbon chart. In explanation we can see that ‘South’ Region is accounted for the majority of decrease among Regions.
We can give negative and positive feedback for visuals and features from icons (thumbs up and down) and by selecting + button we can add selected visual to our report.
Select + button to add Ribbon chart on report:
This was about analysis of Decrease in Sales between Year 2014 to 2015.
Explain the Increase:
Now select datapoint for Year 2016.
We have Explain the Increase option for Analyze. Select ‘Explain the Increase’ option:
Select Sales by Year and Sub Category:
In Analysis we can see ‘Machines’, ‘Copiers’, and ‘’Tables’ had the largest increase among Sub Category. The relative contribution made by ‘Bookcases’, ‘Machines’, ‘Copiers’ changed the most.
By selecting these options, we can find out reason for respective increase or decrease in data points.
Find where this distribution is different:
We can find out where a distribution is different, and get fast, automated, insightful analysis about our dataset. When we select Analyze from datapoint we have one more option ‘Find where this distribution is different’:
Power BI Desktop runs its machine learning (ML) algorithms over the data, and display visual in a new window as column chart with description, and details about values of those category that results in the outstandingly different distribution.
In our case, it will open filters that cause most different distribution of Sales by Year. For State category, it displays that Washington, Florida and Texas states are the most to affect distribution.
In below visual we can see Sales for different years with respect to States.
The visual uses a dual axis to analyze the comparison between the proportion of sales across different Years, for Washington versus different States.
Here we have comparison of overall Sales (grey color) with selected filter applied, in this case Washington (blue color). By default, three different filters are applied for State and Washington State is selected by default.
Different filters can be selected by clicking on them and we can select all filters at one time.
We can give negative and positive feedback for visuals and features from Icons (thumbs up and down) and by selecting + button we can add selected visual to our report.
We can use these insights to analyzing data and to add easily into our reports.