Dividing columns depending on their name

Probably you have changed the workflow (maybe unintentionally).
I suggest you to download the workflow again and replace it with the current one.

Best,
Armin

Ok I reloaded the workflow and it still the same.
But I think that I find out why let me explain :
I have some code with no sales value on the second part of the year.
When I keep the same workflow but applying a row filter node before your part of the workflow in order to keep only some code with some sales value on every column (I mean no 0 sales).
Then it’s working.
But when I have some code with some 0 as sales value then it doesn’t work for all code. Strange isn’t it?

Do you think it could be an explanation?

Are those missing or “0”? I guess missing, because we already have had “0” values in the example and no problems appeared.
If there are missing values you may want to use a “Missing Value” node to convert them to “0” (or some other value that you decide).

Please let me know if this works. (Still I doubt that the only problem is this as missing values in data set will only cause missing values in coverage table.)

Best,
Armin

PS: If you have “0” values for three moths in a row then the coverage value will still be missing as the denominator in the formula is 0.

I already employed a missing value node before in the workflow in order to replace any missing value with a 0. I don’t think that this is the problem.
For example have you tried in your workflow to replace for 1 code sales value & stock value by 0 from 06.2018 till the end of the year and see what happen?

It’s normal to have missing for the stock month for which the corresponding sales months are “0”. The denominator in the formula you provided would be “0” in those cases. E.g. (30 * 190) / ((0+0+0)/3)
You may decide to convert them to “0”.

Best,
Armin

Hi armingrudd, I find out another solution using the moving average node which is working perfectly.
Anyway, thanks for you’re help!

Best regards,
Nicolas

1 Like