I am trying to do a automatic filter to feed a k-means node. I have many numeric features and I want to filter using a correlation filter, but once filter is made I want to connect it by flow variables rigth to column filter node. So the columns that column filter node have to use, they are stablished by correlation filter automatically.
I hope someone can help me. I have dealt with several nodes from Variables (Variables / Flow Control),but it has been impossible.
Best regards friends
would it be possible for your to share your workflow? I am struggling to understand what you are trying to accomplish.
The Correlation Filter node only outputs the remaining colums and they can go straight into a k-Means node, so what is your issue?
In case you are looking for it, the Extract Column Header node can tell you the names of the "surviving" columns.
In the yellow zone of the workflow, I want to use a linear correlation and a correlation filter to feed a column filter before a k-means. But I need to feed automatically that column filter. I was trying with different flow variables nodes but I can´t do it. I wish tou serve this workflow.
I think I understand what you are trying to achieve, which is to pass to the rest of the workflow only the uncorrelated columns out of the correlation filter.
For this purpose, you can use a Reference Column Filter node, using the output of the correlation filter as reference and passing to it the original data table.
I have modified your workflow to include this step (attached).
Thanks a lot. What you did was what I wanted.