Tips for first Clustering in KNIME

[MarcoK]

Hello everybody,

I am a complete newbie in KNIME.

I would like to do an analysis where I split a numeric column into clusters depending on how high the revenue(other column) is. How can I portray such a thing?

Thanks in advance!

Marco

[armingrudd]

Hi Marco,

Would you provide a sample dataset so I can understand what exactly you are trying to do?
As far as I could understand here, if you want to label the rows based on some conditions like:
$Revenue$ < 50 => "Low income"
$Revenue$ > 50 => "High income"
Then you can use “Rule Engine” node.

Rule Engine — NodePit

Best,
Armin

[MarcoK]

Hi Armin,
that’s what I want to do, but I don’t know the ranges (in your example <50;>50). I want to create about 8 “areas”, which are calculated on the basis of the revenue.
$Revenue$ < ? => "Lowest incomegroup"
…
$Revenue$ > ? => "Highest incomegroup"

Then I want to group other columns based on this division.

Best Regards,
Marco

[MH]

You could use the Auto-Binner Node that automatically sub-divides your data based on the selected revenue column and a number bins. You can use equal width option (same range over the value domain) or frequency (same number of samples within each bin) to auto. adjust the ranges, or you can manually specify quantiles as cut off values for the binning.