kumar vishwajeet How is that a 50 dimensional array? It seems to me to be a 2 dimensional array where each row of X could contain one of your X1, X2, etc. vectors. So your X matrix would be 10000 rows by 50 columns. That is a 2D matrix. To get the mean of each row, you'd do
meanOfRows = mean(X, 2); % Take mean of each row going along the 50 columns.
If you want the probability density function of X I'm not sure why you're doing the covariance. Why not just take the hist of the whole thing and normalize: pdfX = hist(X) / numel(X);