SVD/PCA of https://archive.ics.uci.edu/ml/machine-learning-databases/musk/clean1.data.Z

clean1.py

#!/usr/bin/env python

from __future__ import division
import numpy as np

f = open('clean1.data', 'r')
f_nums = [map(float, x.strip().split(',')[2:]) for x in f.readlines()]

d = np.array(f_nums)
d_cols = d.transpose()
means = [np.mean(col) for col in d_cols]

d_mean_centered = []
for row in d.tolist():
    mean_centered_row = []
    for (ety, mu) in zip(row, means):
        mean_centered_row.append(ety - mu)
    d_mean_centered.append(mean_centered_row)
d = np.matrix(d)
d_mean_centered = np.matrix(d_mean_centered)

c = d_mean_centered.transpose() * d_mean_centered #* (1/d.shape[0])