• Replies 3 replies
  • Subscribers 702 subscribers
  • Views 122 views
  • Users 0 members are here
Options
Related

Creating Better YonY "Hocky Stick Charts" with continuous date ranges

Tom Brown (Past Member)
$organization

I'm wondering if anyone has figured out ways to get Tessitura Analytics to produce really good looking continuous sum reports, (adoption curve, or "hocky stick charts)?

Two questions

1. Treating the X Axis as a continuous variable not a order categorical value for line charts.  Note on this chart that -1349 days is showing up prior to the beginning of the campaign is taking way less than horizontal space than the 364 within the current year.  For me a great chart like this gets this detail done in a more visually accurate manner.  Yes, I know that I could likely filter to days >0 and resolve this detail to a great extent.

2. Use real date values on the x axis on this type of stacked hocky stick chart.

Other than using outside tools.  Any ideas?

Thanks for your help.

  • $organization

    You rang?

    If you are using a date, choosing continuous timeline (on your X-axis hamburger menu) will crash your dashboard if you have too many values. It's a known issue (which I'm sad about).  I don't think you can get a continuous 

    For stuff like days prior I'm not aware of a even spacing.  That's on my wish list though

      

  • $organization

    Thanks Heath,

    As you can imagine I hit the bug almost immediately.  How little data is too many values?  Every time I tried I crashed.  I ended having to go to a weekly analysis to cut down the number of records.
     
    I also need relative dates on a fiscal year.  Not Calendar Year dates.
    For now I'm providing 2 graphs. 
    The overlapping one with the confusing relative dates. 
    And the second one that is not overlapped but provides absolute dates.
    I also used the Java Script hack to get the continuous lines.
     
  • $organization in reply to Tom Brown (Past Member)

    I was told for continuous timeline using day as the granularity, that I should stick to a month of data or less - but it's an approximation as it depends on number of data points. 

    For fiscal year "relative dates" is a manipulation of the actual date which is continuous interval and I don't think that you can strip the year out. But for something like Days Since Campaign Start etc there should be javascript for setting the interval. I thought that something like this should work https://support.sisense.com/kb/en/article/change-x-axis-interval but no luck so far