Looping through Columns for multiple predictive models

Hi Folks,

I’m a bit stuck and going around in circles. I’ve got a bunch of stores with various store characterstics. Each store has a bunch of products which either have sales or do not. The store charactersitcs do not change regardless of product, they are global. I’m trying to build a loop to predict sales for each product based on the store characteristics. Previously, when I had a couple of products, I would just build 2 models. I’ve now got hundreds. I managed a bodge solution. I cross joined the store characteristics in SQL with a Product ID / Product Sales column then fed in as a CSV. I could then use the group loop on product ID to do this. However, I want a much slicker solution where I can feed in just one table and get a model for each product column. I’ve enclosed a rough excel of what I’m trying to acheive. If anyone could help I’d be eternally gratefull.

Thanks.
knime.xlsx (11.7 KB)

Hi steven_wynn_uk,

not sure if I unterstand your question correctly.
But it sounds like the node you want is the column loop where you select which columns you want to loop through
Example below:
column loop example.knwf (8.6 KB)

1 Like

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.