normMixEM: EM Algorithm for Mixtures of Independent Multivariate Normals

In srhaup2/clustering_scRNA: Dimensionality Reduction and Clustering For scRNA Data

Description

This is a normMixEM class, which initializes and runs EM algorithm output for mixtures of independent multivariate normal distributions

Arguments

input_data	A matrix/dataframe of size nxp consisting of the data.
num_components	Number of components.

Value

A list including

• convergence - Convergece status. 0 if converged,1 if not.

• mu_mat - A k*p matrix includes the mean parameters of k normals.

• sigma_mat - A k*p matrix includes the variance parameters of k normals.

• pi_vec - A vector of length k includes the probability parameters of k components.

• iter - A integer. Number of iterations

• loglik_list - A number. The log-likelihood of the output result

• prob_mat - A n*k matrix includes the probabilities of being assigned to k components for each sample.

srhaup2/clustering_scRNA documentation built on Dec. 23, 2021, 4:29 a.m.