Exam PL-300 topic 3 question 47 discussion

This question requires "modification" of the scatter plot and what high-density sampling essentially does is to employ methods that capture and represent the underlying data more effectively and eliminates overlapping points. Remember that the table named Data1 contains 10 million rows. How do you represent all that data in a scatter plot in a meaningful pattern for easy understanding and analysis? by use of high density sampling. "By definition, high-density data is sampled to create visualizations reasonably quickly that are responsive to interactivity. Too many data points on a visual can bog it down, and can detract from the visibility of trends". This link explains it more: https://learn.microsoft.com/en-us/power-bi/create-reports/desktop-high-density-scatter-charts#how-high-density-scatter-charts-work

The correct answer should be C - Enable high-density sampling on the scatter plot. Trend line is only available if the X axis is time and type is "Continuous", in this case the option is not available.

I was not aware of this option (and its usage) in the scatter visuals. Initially I thought of option A which would provide more information by using the bubble size in the chart but once I read the artivle shared by shakes103, I chnged my mind. Also in that article it's mentioned: "Too many data points on a visual can slow it down and detract from the visibility of trends" and this can be solved with high density option

According to this site: https://learn.microsoft.com/en-us/power-bi/create-reports/desktop-high-density-sampling it's not available for scatter plots. B should be correct

Don't just rely too much on the wording of documentation, let's consider the question on Statistical sense. The question focus on the EASE of detecting PATTERN, so B - adding a trend line make sense. A trend line visibly show the (linear) relationship between 2 variables: upward - positive, downward - negative, so it help users see the pattern on the spot. And ofc, it's not just for time series data like some other comments has mentioned. Increase higher density of data (C) is basically increase the representation degree of the sample compared to the population. Of course the better representative the sample, the more accurate the trend on the chart. But the point is: this option increase the ACCURACY, not the EASE OF DETECTING PATTERN. Even if the data on scatter chart is more accurately representive, but a bunch of points on it is not easier to show pattern than a trend line, right?

The correct answer should be C. Option B (Add a trend line to the scatter plot) can be more useful when we want to identify a linear relationship or trend between the two variables on the scatter plot. Option C (Enable high-density sampling on the scatter plot) can be more useful when dealing with a large number of data points, as it can help visualize areas of high data concentration. In question, specifically mentioned the data limit 10 M rows so, C is the correct answer.

This is a basic scatter with no Play Axis value. Modifying the scatter is made using Analytics allowing you to add a bunch of trend lines, symmetry shadows and so on. Based on these findings, the answer B looks legit : add trend line(s) such as min, max, average, median, percentile

High density sampling was my first answer, but I think the question is saying that the scatterplot relates to the bar chart and does not contain 10M data points; that would be insane. So, yes, I think the answer is trend line.

This was on the exam on 22/2/2024 (:

The best option for modifying the scatter plot in Power BI to make it easier for users to identify meaningful patterns, without affecting the accuracy of the other visuals, is: B. Add a trend line to the scatter plot. Adding a trend line to the scatter plot in Power BI would enhance the visualization's ability to reveal patterns or trends in the data. This approach is beneficial because it doesn't impact the underlying data used in other visuals, such as the card showing the number of records or the bar chart displaying total transaction amounts by territory. Trend lines provide a visual representation of the general direction of the data, making it easier to identify relationships and patterns without altering the data itself. This approach seems to be the most suitable and voted for, considering the requirements of maintaining the accuracy of other visuals and aiding in the identification of meaningful patterns in the scatter plot.

The keyword in the question is: "10 million rows" => C. Enable high-density sampling on the scatter plot.

Exam: 15/10/2023 Score: 948/1000 Answer: C

For my opinion, B is correct answer. Because the scatter plot that made by grouping territories, so it doesn't have too much points. In this case, the trend line will be more effective for users to understand the pattern of data.

Line chart is not a good choice for a scatter plot. C sounds more rational

https://learn.microsoft.com/en-us/power-bi/create-reports/desktop-high-density-sampling

C. Enable high-density sampling on the scatter plot. Enabling high-density sampling on the scatter plot will reduce the number of points displayed in the visual, while still maintaining the overall patterns and trends in the data. This will make it easier for users to identify meaningful patterns without affecting the accuracy of the other visuals. Adding a count field to the size bucket of the scatter plot or applying a row filter to the Data1 query in Power Query Editor will change the way the data is displayed in the scatter plot and may affect the accuracy of the other visuals. Adding a trend line to the scatter plot may help users identify trends in the data, but it may also obscure some of the underlying patterns in the data. No confusion, no need to discuss further

10 million rows >> high-density sampling