Skip to main content
Announcements
Join us at Qlik Connect for 3 magical days of learning, networking,and inspiration! REGISTER TODAY and save!
cancel
Showing results for 
Search instead for 
Did you mean: 
catalystmichael
Partner - Contributor III
Partner - Contributor III

Trend/Regression Line in a Scatter Chart

I've read through a lot of similar posts on here but can't find a good answer on how to plot a linear trend line in a scatter chart.  E.g. with X as Sum(Sales) and Y as Count(Customer) and Store as the dimension.  

I've used the linest_m and linest_b functions to create a linear trend in a bar/line chart but when it comes to a scatter i can't get it to work.  For the Linest_m function I can add the Y values by creating an aggr() table but it doesn't seem to like it if I try the same for the X values.  

Has anyone had any success with this (using the reference lines)?

@Michael_Tarallo - I've read your posts on using the linest functions in bar and line charts and they make sense.  Any thoughts on how to use them in a scatter where X and Y are both aggregates?

Labels (6)
4 Replies
miguelbraga
Partner - Specialist III
Partner - Specialist III

Hi there,

 

Maybe you could try using a Combo Box type chart instead. I guess the result you need will be the same as a Scatter Chart. Try your approach with a Combo Box.

 

Best regards,

MB

catalystmichael
Partner - Contributor III
Partner - Contributor III
Author

Hi MB,

The problem with a combo chart is that the only dimension is the X axis.  No option to split points/markers out by another dimension like you can in a scatter.

I get where you're coming from in that in a combo chart I could use, for example, aggr(count(customers),store) to create the equivalent x axis of count(customers) in a scatter chart.  The problem still remains of not being able to split out the markers though.  

WernerDC
Creator
Creator

@catalystmichael I also want to know when we will be able to do Regression lines on Scattor plot. It is quite an important feature to have. 

@Micheal Tarallo  The integration with R is not friendly and I cant get it to work. Do you guys have...

pg1234
Partner - Contributor
Partner - Contributor

I am trying to get this to work as well.

I used a trend.qvf post from another thread as a dataset.  My Scatter Plot is based of:

Dimension: Category

x axis: sum(COS)

y axis: sum(Sales)

Based on other threads I read, I can get the m and b properly but I cannot combine them together to get y i.e. y=mx +b.

I tried:

LinEst_m(if(aggr(sum(Sales),CategoryName),aggr(sum(Sales),CategoryName)),aggr(sum(COS),CategoryName))*only({1}(aggr(sum(COS),CategoryName))) +

LinEst_B(aggr(sum(Sales),CategoryName),aggr(sum(COS),CategoryName))

Any help would be appreciated.