Trying to figure out how to do Principal Component Analysis using Covariance matrix. I'm trying to use http://sebastianraschka.com/Articles/2014_pca_step_by_step.html

Now the problem I am facing is that I am not using 2 classes like they have. I have only one set of data with 6000 rows and 784 columns. How can I modify that code to use this input instead of 3-dimensional sample set. I will be using top 10 eigen-vectors. (which might be dimensions?) Basically a single amount of data, unlike their two data sets.

What I've tried seemed to run into complex numbers, and I don't believe that is correct, plus numpy.linalg.eig() takes forever around 10+ minutes.