[SOLVED] Principal Component Analysis in java

Principal Component Analysis in java

I am trying to achieve multiplication in array for doing PCA in java

I calculated mean and substrtacted it from each x values.Next I need to find covarience

So inorder to find that I need to multiply all the combinations in a given array

 [a,b,c] --> (aa)(ab)(ac)(bb)(bc)(cc)

How to construct a matrix of all possible products?

Whether taking subset and multiplying solves the problem?

Solution

You are computing a matricial product. Say A = [a, b, c] (horizontal vector), you get an obviously symetric matrix by :

M = ^tA . A

The upper part of the matrix is composed of all possible products.

aa ab ac
ba bb bc
ca cb cc

When computing, you can use the symetry :

for(int i=0; i<len; i++) {
    for (int j=0; j<=i; j++) { // do not go up to len but stop at i ...
        // computations ...
    }
}