I have an input set that has been sampled and normalized. I now want to divide this dataset into 2 subsets based on the value of one attribute only. eg: The feature is 'has_children'. I want to segregate the instances depending on has_children=0/1 and then apply clustering to each subset.
I am not trying to create training-test sets for validation. If anyone has any pointers, please let me know.
you can use the Generate Attributes operator under Blender->Attributes->Generation.
Then you can work with a if rule and assign a numeric value 1 if the string takes a value "has_children" and other in other case.