'Does a plug-in selector bivariate kernel density estimator with weights exist for python?
I am trying to calculate the kernel density estimate for a set of weighted bivariate data points. I am currently using KDEpy.FFTKDE. However, this does not provide a method for automatically calculating the bandwidth. Do you know if there exists a multivariate KDE estimator which allows the data to be weighted and uses a plug-in method to calculate the bandwidth for python?
I typically have about 10^6 data points so I believe cross-validation will be too computationally slow.
fastkde nearly provides what I want but doesn't allow the data to be weighted.
I believe the kde function inside of the sm package in R provides what I want (see page 33 of this). However, I've never programmed in R.
Solution 1:[1]
You may try scipy.stats.gaussian_kde but the speed of computing will be slower compare to KDEpy.FFTKDE.
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|---|
| Solution 1 | user19147325 |