1 Reply Latest reply: Jul 24, 2017 10:49 AM by Fabien Gaiffe RSS

    curve fitting / best fit curve - least squares method ?

    Fabien Gaiffe

      Hi dear community of Qlikers !

       

      You'll find enclosed the excel table I'm referring for this issue, and the qlik sense app related.

       

      So ! As you will see, I have a quite simple table with three raws : Product type, Year and Sales. I draw on Qlik Sense the visualization showing the evolution of the sales per year and per Product type. So far, so good.

      If you look carefully at the data, I don't have any Sales figure for product type D, in 2010. It's intentionnal. As we can see, the red curve of product D is quite similar to the blue curve of product A, but way different of the profiles of the curves for product B and C.

       

      What i would like to do is finding a way in Qlik Sense to find which Product type has had the most similar Sales evolution to my product D (in our case, we can see it, it's product A). Secondly, if this is possible, I would like to "forecast" the evolution of product D, based on the evolution of this most similar curve.

      I guess I will have to use like a least squares method or something, but I really don't figure out how to perform it...

       

      I hope I have been clear in my description ! Don't hesitate to ask me more precisions if not.

       

      And thanks so much in advance for your help.

       

      Long life to Qlik comunity and the cooperation it brings !!