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
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!
Hi Paul, thanks for ur reply but i believe the formula is returning the wrong figures ... My data list has more then 30000 Tests and the formula is returning 22000 which im pretty certain is not the correct figure.
No need for a column for “rows”. I only included it to make it easier to check that the answer is correct.
How do you know it isn’t correct?
EDIT: reading through your posts, you first mention yor dataset has over 30,000 tests, and then say the table lists more than 150,000 rows. Is each row unique? (That is, does each row reflect a unique test?)
Sorry i forgot to mention that the tests are not unique on my data files as they are at least tested twice a year across different locations. I am of course using mapping tables for the periods and the regions and its so far working for me.
Can you please advise me on how to proceeed with duplicate controls ?
And i know the number is smaller because according to my predecessor, the number is alwas less than 5% of tests not performed.