how to create a mixture of gaussian distributions from data points generated using normal distribution

Question

I'm trying to create a mixture of gaussian distributions in 2D, 5D and 10D from data points that follow normal distribution. Till now i was able to create data points that follow normal distribution using.

import numpy as np
data = np.random.normal(0.1, 0.3, size=(5000))

My question is how can I create mixture of gaussian distributions (in 2D, 5D, 10D) from the above data points ?

Answer 1

It depends on whether you wanna to create random variable where of individual distributions are independent or there is some covariation between distributions.

If there is presence of covariation matrix, I will use vector of uniform distributed random variables, and transform this vector into vector of random variables with normal distribution using covariance matrix.