How to resample continuous random variables based on a 2D probability density function

I have two-dimensional data and n number of data. I can have the Kernel density estimation of this data, how can I generate more data based on this distribution?

Answer 1

[Arun Mathamkode]

I believe the randsample function will be helpful to you. You can use the density estimate from the kernel density estimation as the weights in randsample. You can also use Metropolis-Hastings sample method ( mhsample ) but it may be more suited if you have an analytical expression of the pdf.