MATLAB實例：二元高斯分布圖

作者：凱魯嘎吉 - 博客園 http://www.cnblogs.com/kailugaji/

1. MATLAB程序

%% demo Multivariate Normal Distribution
clear
clc

%% 空間坐標范圍
x1=-5:0.2:5;
x2=-5:0.2:5;
[X1, X2]=meshgrid(x1, x2);
X=[X1(:) X2(:)];

%% 高斯分布參數
% 分量1
miu_1=[1 1]; % 均值
Sigma_1=[2 -2;-2 3]; % 協方差
% 分量2
miu_2=[0 -2]; % 均值
Sigma_2=[5 0.5;0.5 1]; % 協方差

%% 高斯概率密度函數
% 分量1
y_1=mvnpdf(X, miu_1, Sigma_1);
y_1=reshape(y_1, length(x2), length(x1));
% 分量2
y_2=mvnpdf(X, miu_2, Sigma_2);
y_2=reshape(y_2, length(x2), length(x1));

%% 2D密度圖
figure(1);
contour(x1, x2, y_1);
hold on
contour(x1, x2, y_2);
xlabel('x1');
ylabel('x2');
saveas(gcf,sprintf('Gauss_2D.jpg'),'bmp');

%% 3D密度圖
figure(2);
surf(x1, x2, y_1);
hold on
surf(x1, x2, y_2);
xlabel('x1');
ylabel('x2');
zlabel('Probability Density');
saveas(gcf,sprintf('Gauss_3D.jpg'),'bmp');

2. 結果

參考：Multivariate Normal Distribution