I have a very large data frame (500K rows, 50 columns), for which I want to calculate the density estimate.
Since pdfCluster cannot handle this kind of dimensions, I was wondering if there is a way to calculate the density based on a sample of, say, 10K points and predict on the other 490K.
However the result of pdfCluster just gives the point estimate at the 10K points, and I can't find an underlying function/object from which to extrapolate on the new points.
Is there a way to do that?