Forum Discussion

Mous007's avatar
Mous007
Icon for Helper IV rankHelper IV
6 years ago
Solved

New measure or column ?

Hello everyone,  Here is a little extract of my data:   Tests Validity 1 Validity 2 Facility owner approval Final validity Results Test 1 Valid Valid Yes Valid ...
  • PaulDBrown's avatar
    PaulDBrown
    6 years ago

    Mous007 

     

    Ok, I have a feeling that the measures are correct, but I suspect you are getting the results you see because, given the 150,000 + rows of tests, and given that you carry out 230 individual tests, if you analyse the whole dataset you are finding that all tests have been carried out at some stage, and none of the tests have never "not been carried out". 

    150,000 rows / 230 tests = 652 rows / test (on average)!!

    That is, If I am understanding what your are trying to achieve (as per your original post).

     

    What I suggest you try is to add a slicer to your page (year for example) and play around to see if the values change. 

    I have tried it in my sample and the measures return what I am expecting. I have changed one of the measures to the below to allow for the slicer to work (changed the ALL in the filter to ALLSELECTED):

    Unique Tests Not Executed EVER = 
    VAR calc = CALCULATETABLE(VALUES(Sales[Product]); FILTER(ALLSELECTED(Sales); [Not Executed (at some stage)]=0))
    VAR full = VALUES(Sales[Product])
    Return COUNTROWS( EXCEPT(full; calc))

    Here is an example if I filter my sample data to see only rows for 2018 (which you can compare to the image I posted in the above post):

     

     

     

    But there again I may have misundestood what you are actually trying to achieve!