How can I get mean and covariance of built GMM from samples? about gaussian-mixture-model HOT 1 CLOSED

Dear John,

thank you for the interest in this library.
You can build the GMM from samples using for example expectation maximization:
gauss::gmm::GaussianMixtureModel learnt_model(
gauss::gmm::ExpectationMaximization(samples, number_of_expected_clusters)
);

Then, you can always access the set of clusters composing the GMM calling GaussianMixtureModel::getClusters().
Each element in the returned vector is a gauss::gmm::Cluster structure, containing the weight as well as the Gaussian distribution associated to a particular cluster.
You can access the mean or covariance of such a distribution by calling GaussianDistribution::getMean() and GaussianDistribution::getCovariance().

from gaussian-mixture-model.