Hm I don't fully get it... You are talking about columns, but the aggregation is only over rows?
In your second example: In line 3 it was the first time A showed up. Why is it 2 times in 3 events?
In your first example: Are both columns one set? Why is the value of the third row in col2 33? shouldn't it be 66? And the fourth row in col1 why is it 75, I would have thought it is 25?
I will attach you a workflow which solves the problem... at least in the way I understood it.
I think I now understand. You are counting te percentage of x after each row. So down the column.
my quick thoughts are that it's not easy. I would first use the one2many node to get a count of 0 or 1 in an an X column. You would then need to generate a row index using the maths formula node (row index+1).
now the tricky part is accumulating the values in the x column. You will need to do this using the delegating loop start And end nodes, but these are nodes that need some good knime understanding.
finally you would take your accumulated values and divide by the row index column multiplied by 100.