Split function: what is happening when one column is not added into "Group"

[#1]

I got a problem with SPLIT when 1 column was not included in Group. I intended to make it not in Group and hope it the values in another column can be averaged.

So I made a sample Data Table to play and try to understand what's going on.

Sample table below:

Scenario1:

I want to Split "Random metric value", Split by Tool A&B, Group by only "Test", and hope the "random metric value" from different iterations can be averaged for the same "Test".

Here is what I got:

Scenario 2:

But when I do manual averaging (Summary-Mean(Random Metric Value) then split by tool A&B), the results are different:

Above should be the right results.

But my question is: what is happening when I do SPLIT in the 1st scenario without cast "Iteration" column anywhere?

Best regards,

Zihao

[#2 · Accepted solution]

The split does not do any calculations.  In your first case, it is selecting the last iteration value for each test and placing that into the appropriate Tool A or Tool B.