how to use recursive and chunk loop in the same time

Hello,
I want to design a loop that goes through row #1 and outputs a value that will be used in row #2. So it has to go through the rows one by one (not all at once) and bring back the data from the loop end node to the start node. So basically a combination of chunk and recursive loops. Is there a way to do this? Any ideas you may have are appreciated!

Hi @lamasewidan , could you share some data and show what the expected results for the first 3 lines, and of course explain how you got to these values?

This will help illustrate what you are trying to do, and there might be a more straight forward way to do this instead of using 2 loops nested in one another.

Thanks for the reply! Okay this is what I want to do
Capture
This is what the initial data would look like.
I want to look ONLY at row #1 and do the following: Stock-orders and the result would be identified by “gap” (maybe a new column named gap).
And then we move to row #2, if the gap of row#1 is >=0 subtract the orders of row1 from the stock of row 2 and this would be my new stock for row#2, else keep the stock of row#2 as it is
then calculate the gap for row#2 like we did for row#1, stock-orders (stock of row#2-orders of row#2 ofcourse)
and then we move to row#3
if the gap of row#2 is >=0 subtract the orders of row2 from the stock of row 3 and this would be my new stock for row#3, else keep the stock of row#3 as it is
then calculate the gap for row#3 like we did for row#2 and 1, stock-orders (stock of row#3-orders of row#3 ofcourse)
and then we keep doing the same thing if there are more rows.
I had 2 issues with doing this logic in knime; the first is going through every row individually which only the chunk loop does, the second is bringing back the gap of a row to be used in the math of the row after it and this can only be achieved by the recursive loop
thats why I initially asked how to used the 2 loops together. if there are other ways to do this please let me know!

@bruno29a check reply above

I guess Nodes like Math Formula, Moving Aggregation or Lag could be helful here
br

Furthermore please check these posts they seems very comprehensive and detailed (more complex what you need but maybe they will give you some inspiration) I am sure you will find what you are looking for:

Kind regards!