I am somewhat a beginner to KNIME, so I am sorry if this question is covered somewhere in the documentation.
I know how to set a fixed number of inputs and outputs for a new node. However, I want to create a node which can theoretically take in any number of inputs rather than a predefined set of inputs. The upper limit could be in the thousands of inputs, though most likely it would be 5-50 inputs. Is there an easy way to do this?
Short answer: no. Long answer: first you have to define the number of ports statically during development (though there are optional input ports, which add some kind of dynamics) and second a node cannot sensibly have more than four or five ports, they simply don't fit in the workflow editor.
What use case do you have in mind for such a large number of input ports?
It is an ensemble method of sorts. The node would take in k number of different classifier outputs for each sample, along with the original sample features, and then learn how to weight the value of each classifier output to combine them into a prediction for a new sample.