Forum Discussion

FrankWB's avatar
FrankWB
New Member
3 years ago

Calculation groups or Unpivot

Hi,

I have a data set of monthly Stock data. It is a flat table of about 250,000 lines.  Every line contains, besides the basics: date, stockID, price, 100 featurs of the stock. Things like Pricer/Earing ratio, market cap, etc.

My task is to create measures such as TopN based summations based on these features.

I have two choices (maybe more):
1)  Unpivot the data and then have 25 mio lines with two columns feature_name and feature_value

2)  Create a placeholder measure for every feature abd the use calculation groups (in the Tabular Editor) to create the requiered measures on SELECTEDMEASURE.

Which one would you chose and why ? 
I'm particular worried about performance issues. Also the restrictions of using calculation groups.

Many thanks in advance

 

"Calculation Group"
unpivot

1 Reply

  • some_bih's avatar
    some_bih
    Community Champion
    3 years ago

    Hi FrankWB currently there is "testing period" for DAX optimization for free, check link below. Join and hopefully you will have chance to test solution / get answer about your model for free. I am not sure is paid version already working.

    I would say this is great example to be candidate to get "optimal" answer to your question. Hope this help in finding your solution.

    https://www.daxoptimizer.com/