MORNING,

Please, I'm newer on KNIME.

I would ask something about distance matrix.

I have a SQL tABLE of of customers (200,000) which are described with qualitative variable.

So to avoid memory problem, I transform all these qualitative variable into numeric (i.e, 1 for good, 2 for medium, 3 for bad, NO order relation) and each customer is described with 17 variables.

I want to do clustering now on my table. So What algorithms I have to use to cluster my table with Categorical value?

I thought about distance matrix but When I connect my table to "distance matric calculate" node and I choose "dice" (I can't choose euclidian, because it is qualitative variable) distance for example, all my columns disappear. even I transform my numerical values to string (because it is qualitative before) , nothing happen

Please help, do you know a knime methode for clustering on categorical variables?

How do you use "distance matric calculate" node?

Thanks