Creating a Custom Segment of a Clustered Bar Chart

User_32612
User_32612 Member
in Magic ETL

Hello,

With the recent update which added segments I was wondering if it is now possible to add custom segments of a clustered bar chart. For example, with the chart below I have YoY sales by week, but filtered to only show week 16-28:

image.png

I see in the segments tool that it is possible to create segments which ignore filters, would that make it so I can create a custom segment, in this case something along the lines of a "pre week 16" average set of bars to go before WK16?


On a side note, is there any way to remove the spaces between the bars within the same category in the below chart so they look similar to the clustered + symbol chart above without having to use symbols?

image.png

Thank you!

Answers

  • jaeW_at_Onyx
    jaeW_at_Onyx Coach

    handful of questions. why 16 to 28? is that b'c you're specifically interested in that point of time or is the timeframe dependent on today()'s date? (in which case build your filters and axis on a variant of the current_date() function.


    if it were me I would build an offset date dimension table so I can use other chart types than just the Period Over Period. https://www.youtube.com/watch?v=CDKNOmKClms&t=142s

    if your months are selected based on a lag from today's date, i'd implement that in the offset date dim.

    Jae Wilson
    Check out my 🎥 Domo Training YouTube Channel 👨‍💻

    **Say "Thanks" by clicking the ❤️ in the post that helped you.
    **Please mark the post that solves your problem by clicking on "Accept as Solution"
  • User_32612
    User_32612 Member

    @jaeW_at_Onyx The time frame is defined by a beast mode which filters out any week after 15 and any week prior to the current date.


    What I'm looking to do here is essentially what you see in this excel chart below (ignore the # & axis differences, this is separate data):

    image.png

    The reason we want to go this route is the chart doesn't have too many bars that it isn't easily readable.

    Is the tutorial you linked capable of producing the end result seen in this excel chart?

  • User_32612
    User_32612 Member
    image.png

    Essentially what I am asking is is there a way to aggregate columns that have the same name and average their values into one column? As you can see above, when I use a beast mode to rename weeks 1-15 as Pre WK16 AVG, instead of aggregating them they are each shown.

  • GrantSmith
    GrantSmith Coach

    @User_32612

    I'm assuming you're using a beast mode for your x-axis - do you have any fields listed in your sorting? That may throw off your grouping. You may want to come up with a similar beast mode to define your sorting such that anything pre week 16 would be a value of 15 so that they'd sort and group based on the same value.

    **Was this post helpful? Click Agree or Like below**
    **Did this solve your problem? Accept it as a solution!**
  • User_32612
    User_32612 Member

    @GrantSmith

    I decided to make a calculated column in my dataflow and that successfully aggregated the "Pre WK16 AVG" into one column. I hadn't realized that the calculated columns in data flows behave differently that beast mode, this should be useful info moving forward. Thanks for the assistance!

  • GrantSmith
    GrantSmith Coach

    Glad you got it figured you!

    **Was this post helpful? Click Agree or Like below**
    **Did this solve your problem? Accept it as a solution!**

Categories