i'm using Knime 3.1 on Windows10, and i want to show you an issue that i have with the node "Auto-Binner" in combination with "Row Splitter". I don't know if it's a bug or a wrong procedure...
I try to explain, i have this sequence:
Node 1: an initial table with a numeric integer column.
Node 2: a row splitter to split out a high outlier value (just one row in this case).
Node 3: on the filtered table (with no outliers) i apply an auto-binner to obtain 3 bins of equal width.
I expected 3 different bins based on equal width, but in the output table i can see only one bin.
After a test i notice that the auto binner works well only if i manually delete the row with the outlier in the first table.
It seems that after the row filter the outlier value is still present (in memory? I don't know..) so the binning is wrong.
I attach here an example workflow, so you can see the problem.
Please, let me know.
thank you in advance!
Ps. Compliments for Knime software and great forum/support!