Scatter chart with non-numeric category values

Question

Summary

Content

Hi,

We have a use case where we need to show a scatter chart with non-numeric category values. For example, we need to display the chart in the following manner :

But this is how these values are plotted with BI Publisher:

The category labels are not proper. Is this because scatter charts won't support non-numeric values or are we missing something here.

Thanks,
Deepa

Thomas Dodds · Answer

Why the insistence on scatter chart?  You are trying to convey a conclusion about the data -- that conclusion should be readily visible in the chart/graph/diagram.  The Scatter answers a very specific question about what is plotted.  Mimicking a scatter (without using the fundamentals of a scatter) is nothing more than confusion.  Who is pushing the scatter?  Why?  Why hasn't the fundamental principles of the scatter been explained?

Given your ask (so far) I don't see you driving at relationship or distribution between two variables (measures) ...

Here's a readily findable infographic on selecting the appropriate visual:

User_ZSQFX · Answer

Hi Thomas,

Thanks for the reply.

So does that mean that scatter charts support only numeric values? And if yes, is there any way to customize some other charts so that we can come up with a similar appearance as that of scatter charts?

Thanks,

Deepa

Thomas Dodds · Answer

A scatter plot by definition shows the relationship (if any) between two measures ...

What you are missing:  you have a requirement to show a scatter plot ... how did that happen?  Let your users dictate WHAT not HOW.  WHAT is it they want to analyze? You, as the BI expert, then choose HOW best to achieve that.

Given what I see above, a scatter is the wrong visualization -- you only have one measure (count of SR) you are working with.  Had you had two measures (count of SR, avg duration of SR) you could use this type of visualization to see if there is a relationship between count and duration across dimensions like status, agent, day-of-the-week, etc.