Why does Regression-predictor needs training-data column??

KNIME Analytics Platform
1

Hi,

I am using the 'Polynomial Regression Learner' (PRL)and the 'Regression predictor' (RP) nodes. I use the experimental data as input for the PRL node. Then I want to use the generated model to PREDICT these data for a new set of input points. However, the RP node complains that the experimental data column is not present in the new input data. Yes, of course, since I want to predict that data, right? This is the whole point of the prediction. So, is that a bug or why does the RP node complain about that? I use the exp. data only as target column not as independent variable!

best regards

Soren

2

The regression predictor certainly does not need the target/dependent variable column. Can you double check the error message and the columns provided to the learner and predictor node?

3