Kernel density estimation in python for points with masses

by Igor Markov   Last Updated September 07, 2018 05:19 AM

When using sklearn.neighbors.kde, I'd like to specify a mass for each point. For example, mass 2.0 would be equivalent to using two copies of the point, but masses can be large and don't have to be integer. Am I missing something? Should I look into different packages? (sklearn is attractive because it supports approximation and scales well too large multidimensional pointsets)

Related Questions

Sampling from Gaussian Process Posterior

Updated June 18, 2018 10:19 AM

Obtaining SVD eigenvectors on kernel PCA and sparse PCA

Updated November 05, 2017 00:19 AM

Gamma as inverse of the variance of RBF kernel

Updated December 06, 2017 13:19 PM

How to represent a non-trivial kernel as a Gram matrix?

Updated September 09, 2017 14:19 PM