Issue with MAQ Boxplot - Outliers Disappearing

[lucianolorenti]

Hi everyone,

I'm currently facing an issue with the Box and Whiskers by MAQ software and would appreciate some help.

I have a large dataset that includes outliers. However, when visualizing the data using the MAQ boxplot, it seems like there's subsampling occurring. This becomes apparent as some outliers disappear depending on the range of data I use. When I use a smaller subset of the data, the visualization appears correct, suggesting that there might be no subsampling.

I've have a pbix example with synthetic data. Towards the end of the dataset, there's an outlier with a value of 2000. If I visualize the entire range of data, that point becomes invisible. Yet, if I reduce the range, the outlier suddenly appears.

Has anyone encountered a similar issue, and is there a workaround or solution for this?

Dashboard that shows the problem:

https://engit-my.sharepoint.com/:u:/r/personal/lucianorolando_lorenti_indx_com/Documents/BoxPlotExam...

Thanks in advance for your help!

[Anonymous · Accepted solution]

Hi @lucianolorenti ,

Thanks for the feedback, I did some research to investigate. This could be due to the way the box plot is calculated and the way the outliers are determined. But unfortunately, I was not able to find the most recent calculation and the older connections are outdated.Solved: Re: Box and Whisker chart by MAQ Software incorre... - Microsoft Fabric Community

What it says is that outliers in a boxplot are typically defined as values that fall below Q1 - 1.5IQR or above Q3 + 1.5IQR, where Q1 and Q3 are the first and third quartiles, and IQR is the interquartile range (Q3-Q1). If your data doesn't have any values that fall in these ranges, then no outliers will be shown.

For any further queries or enhancement requests, you can contact at support@maqsoftware.com

Hope it helps!

Best regards,
Community Support Team_ Scott Chang

If this post helps then please consider Accept it as the solution to help the other members find it more quickly.