df = pd.read_csv('data/x_train.csv', sep=';', header=None)
c = df.ix[:,:].corr()
np.where(c > 0.95)