My data consists of 400000 samples with 3200 values of same units (400000x3200). I know, when the data has values of different size, first you have to standardize or normalize the data, so the scale of each value is the same.
But how to calculate PCA of data, which has values of the same units? Is it necessary to normalize, standardize, center the data before? What about the cov-matrix for values with same units?
I hope somebody can help. Thanks