Nth degree polynomial trend line support in Qlik Sense 2.1.1?

[#1]

It appears that via the linest_m and linest_b functions, you can generate a first degree trend line of the form y=mx+b in Qlik Sense, but there does not appear to be any support for generating higher order or nth-degree polynomial lines like there are in QlikView.

Is there a work-around for generating 2nd order trends?

Is there a plan for adding first-class support for n-th degree polynomials?  Even for linear trends it's quite awkward to use linest_m and linest_b to add a simple trend line.

[Gysbert_Wassenaar]

You can add polynomial trendlines using the Trendlines option on the Expressions tab. But there are no functions to manually create such trendlines like you can with the linest_* functions. Considering that Qlik Sense is a BI application and not a Statistics application I don't expect all the exotica of a staticians toolbox to be added. Perhaps someone will find a way to marshal data to and from R in an easy and fast way.

[#3]

> You can add polynomial trendlines using the Trendlines option on the Expressions tab.

I should have been more specific in that I was asking about Qlik Sense 2.1.1.  I don't see an expressions tab or an option anywhere to add a trend line and "trendline" doesn't return any hits when I search the documentation.  I would appreciate a pointer if I missed it.

> Considering that Qlik Sense is a BI application and not a Statistics application I don't expect all the exotica of a staticians toolbox to be added. Perhaps someone will find a way to marshal data to and from R in an easy and fast way.

If this functionality is available in QlikView, I would expect it in QlikSense.  This isn't an esoteric function.  This isn't n-variable linear regression where you need matrix operations.  Polynomial trendlines are available in Excel out-of-the-box.  Furthermore, the whole reason for doing trendlines in the first place is to extrapolate out future behavior.

I can't imagine a more basic BI function than "at our current rate of growth, when will we hit $X in sales?".  Growth is almost always accelerating or deaccelerating so a first degree polynomial (e.g. y = mx+b) trendline will almost always be a poor fit for the data.

[Gysbert_Wassenaar]

> You can add polynomial trendlines using the Trendlines option on the Expressions tab.

Ah, my mistake. That only works in Qlikview. Qlik Sense lacks this functionality at this time.

I can't imagine a more basic BI function than "at our current rate of growth, when will we hit $X in sales?". 

I can. Count things. The Sumerians began doing this basic BI thing around 3000 BC.

In my opinion polynomial equations are not basic BI. Perhaps Excel is a better fit for your BI needs.

[richbyard]

I have been looking at the expressions for trend lines and have documented at Calculating trend lines, values and formulas on charts and tables in Qlik Sense.

However, I have yet to work out the polynomial ones so will be working on this over the next week or so... I'll keep you updated if I manage to solve it...

[Gysbert_Wassenaar]

Perhaps using a Vizlib extension that has trendlines built in is an easier solution: Vizlib following the Trend