Counting or Grouping?

Hello, embarrassed to ask what's probably a fairly basic question… but hey, we're all new at some point, right? So here goes:

I'm building an app for a department at my org to bring in event data from an external platform (API hurdle crossed✅). I would like to show a few cards for things like an event count per month, per year, YOY, etc. in addition to views by number of registrants in the same categories. Unfortunately, I'm having trouble conceptualizing how to count unique event IDs instead of registration IDs.

Perhaps this needs to be a grouping tile within the ETL. Is this possible while keeping all the data going to a singular output? Do I need to have a different output? Trying to be conscientious of upcoming consumption-model changes and conserve credit utilization once we transition. Happy to provide screenshots of anything needed.

Is it a better practice to try and achieve this within a Beast Mode? If so, I'm happy to create a new thread in that product category.

I recently switched from a design/marketing role to one where I'm trying to learn some data science on the fly - so I appreciate any patience in potential knowledge gaps. Open to recommendations on how to get the experience needed to skill up and understand these basic concepts. 🙏

Best Answer

  • MarkSnodgrass
    Answer ✓

    @Domolex welcome to the forum (and there are no bad questions here). It is certainly smart to first try and build what you need off the source dataset and see if you can skip the ETL execution.

    To count distinct event ids you can create a beast mode that looks like this:

    COUNT(DISTINCT eventid)

    You can put your date on the x-axis and this beast mode on the y-axis and set the date range filter to graph by month and you would have your events by month.

    **Check out my Domo Tips & Tricks Videos

    **Make sure to <3 any users posts that helped you.
    **Please mark as accepted the ones who solved your issue.

Answers

  • MarkSnodgrass
    Answer ✓

    @Domolex welcome to the forum (and there are no bad questions here). It is certainly smart to first try and build what you need off the source dataset and see if you can skip the ETL execution.

    To count distinct event ids you can create a beast mode that looks like this:

    COUNT(DISTINCT eventid)

    You can put your date on the x-axis and this beast mode on the y-axis and set the date range filter to graph by month and you would have your events by month.

    **Check out my Domo Tips & Tricks Videos

    **Make sure to <3 any users posts that helped you.
    **Please mark as accepted the ones who solved your issue.
  • Thank you, sir! I'll take another stab at this on Monday and be sure to <3 your response or ask more questions.