hi, I am also experiencing some weird behaviour us...
# gd-beginnersj
Jamie Lawrie
10/26/2023, 1:20 AMhi, I am also experiencing some weird behaviour using show as % in visuals - see thread
Jamie Lawrie
10/26/2023, 1:22 AMI have a donut chart showing raw numbers
Jamie Lawrie
10/26/2023, 1:22 AMI’d prefer to present that with %s
Jamie Lawrie
10/26/2023, 1:22 AMIf I use show in %
Jamie Lawrie
10/26/2023, 1:22 AMI get nonsensical numbers
Jamie Lawrie
10/26/2023, 1:23 AMthe donut should obviously add up to 100% - why is that segment showing 95% when it looks like ~30%
j
Jakub Kopecky
10/26/2023, 7:36 AMhey Jamie, are the types unique? is it possible to break down the report and verify a tiny part to make sure the source data are correct?
Jakub Kopecky
10/26/2023, 7:38 AMthis applies to the previous question too I believe it will be connected to the same root cause so I would suggest to start with the numbers and filters and once we understand the break down the % mystery will be uncovered too
j
Jamie Lawrie
10/26/2023, 6:45 PM@Jakub Kopecky I’m going to do testing today so will dig in a little further but yes every fact table record has a type and there are ~10 types possible as you can see from the graph. This is pretty straight-forward stuff - I’m not sure why clicking the show as % button would break it though?
Jamie Lawrie
10/26/2023, 8:15 PM@Jakub Kopecky found the issue - was a join creating unexpected duplicate IDs when I was expecting none. Thanks for the prompt to check
👍 1
i
Ivana Gasparekova
10/26/2023, 8:18 PMHi @Jamie Lawrie,
Thanks a lot for sharing your findings with us.
I am glad you found the root cause of the discrepancies!
👍 1
j
Jan Soubusta
10/27/2023, 3:57 PMWe are going to release a new feature called "Data profiling". If you define a primary key, it will warn you in LDM Modeler app about duplicates.