Hi @Chenfang , if you used @berserkersap 's idea/solution, then perhaps you should mark @berserkersap 's post as solution, not your post ![]()
I’ve created something more dynamic that will create Qx (Q1, Q2, … Qn) dynamically at every 3 columns, and each of the Q columns contains the sum of the 3 columns, so it does not matter if your table has 3 columns or 39 columns or 300 columns, it will work the same way. It also does not matter what the column names are, it will work the same way.
This is what the workflow looks like:
Input data of 12 columns (same as in your original post):
Results:
And now I plug another sample data to the workflow:
This sample data has 39 columns with different names from my previous sample:
Results (it added 14 Q columns, from Q1 to Q14):
I did not need to change anything in the workflow. Just the data changed, and it worked the same for both cases.
Here’s the workflow: Create Q columns every 3 columns containing their sum.knwf (37.4 KB)



