gaussian_mixture_model.h

#ifndef GAUSSIAN_MIXTURE_MODEL_H
#define GAUSSIAN_MIXTURE_MODEL_H
 
#include <Eigen/Dense>
#include <Eigen/Sparse>
#include <vector>
 
 
class GaussianMixtureModel
{
    private:
        int k_;                                 // number of components
        int dim_;                               // dimension of data
        std::vector<Eigen::VectorXd> mu_;       // mean vectors
        std::vector<Eigen::MatrixXd> sigma_;    // covariance matrices
        std::vector<double> w_;                 // weights of components
        double log_likelihood_;                 // log likelihood of data
        
 
    public:
        GaussianMixtureModel(Eigen::MatrixXd samples, int num_components);
 
        GaussianMixtureModel(std::vector<Eigen::VectorXd> means, std::vector<Eigen::MatrixXd> covariances, std::vector<double> weights);
 
        GaussianMixtureModel(int num_components);                                                                    // constructor: create GMM with random means and covariances    
 
        GaussianMixtureModel();
        
        ~GaussianMixtureModel();        
        
        void fitgmm(Eigen::MatrixXd samples, int num_components, int max_iterations, double tolerance, bool verbose);                     // expectation maximization algorithm
 
        void fitgmm2(Eigen::MatrixXd samples, int num_components, int max_iterations, double tolerance, bool verbose);                     // expectation maximization algorithm
 
        void fitgmm(std::vector<Eigen::VectorXd> samples, int num_components, int max_iterations, double tolerance, bool verbose);                     // expectation maximization algorithm
        
        std::vector<Eigen::VectorXd> getMeans();
 
        std::vector<Eigen::MatrixXd> getCovariances();
 
        std::vector<double> getWeights();
 
        double getLogLikelihood();
 
        void setMeans(std::vector<Eigen::VectorXd> means);
 
        void setCovariances(std::vector<Eigen::MatrixXd> covariances);
 
        void setWeights(std::vector<double> weights);
 
        bool check();
 
 
 
 
        double gauss_pdf_2d(Eigen::VectorXd q, Eigen::VectorXd mean, Eigen::MatrixXd cov);
 
};
 
#endif // GAUSSIAN_MIXTURE_MODEL_H