1
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

More than 5 years have passed since last update.

Python - 数値データファイルからデータを読み込み分散共分散行列、固有値、固有ベクトルを求める

Posted at

以下のようなタブ区切りの数値データ系列ファイルからデータを読み込み分散共分散行列、固有値、固有ベクトルを求めるプログラム。

MultipleRegressionAnalysis_Data2.txt
23.3	19.8	14.7	19.7	16.9	14.4
5.24	-5.23	-7.95	11.70	-2.44	-2.14
MultipleRegressionAnalysis_3.py
import numpy as np
import math

def load_data(filename):
    with open(filename) as lines:
        return [[float(data) for data in line.strip().split('\t')]
                for line in lines]
    
def calc_sum_of_square(data_list1, data_list2):
    return (len(data_list1) * sum([data1 * data2 for data1, data2 in zip(data_list1, data_list2)]) - sum(data_list1) * sum(data_list2))/(len(data_list1)*(len(data_list1)-1))

def calc_variance_covariance_matrix(data_list):
    return [[calc_sum_of_square(data_list1, data_list2)/(len(data_list1)-1) for data_list1 in data_list]
            for data_list2 in data_list]
            
def calc_engenvalue_2d(mat):
    lambda1 = ((mat[0][0] + mat[1][1]) + math.sqrt((mat[0][0] - mat[1][1])**2 + 4*(mat[0][1]**2))) / 2
    lambda2 = ((mat[0][0] + mat[1][1]) - math.sqrt((mat[0][0] - mat[1][1])**2 + 4*(mat[0][1]**2))) / 2
    return [lambda1,lambda2]
    
def calc_engenvector_2d(mat, engenvalue):
    a11 = math.fabs(mat[0][1]) / math.sqrt((engenvalue[0] - mat[0][0])**2 + mat[0][1]**2)
    a12 = (engenvalue[0] - mat[0][0])*a11 / mat[0][1]
    a21 = math.fabs(mat[1][0]) / math.sqrt((engenvalue[1] - mat[0][0])**2 + mat[0][1]**2)
    a22 = (engenvalue[1] - mat[0][0])*a21 / mat[0][1]
    return [[a11, a12],[a21, a22]]
            
if __name__ == "__main__":
    data = load_data('MultipleRegressionAnalysis_Data2.txt')
    variance_coveriance_matrix = calc_variance_covariance_matrix(data)
    print("- 分散共分散行列 -")
    print(np.array(variance_coveriance_matrix))
    eigenvalue = calc_engenvalue_2d(variance_coveriance_matrix)
    print("- 固有値 -")
    print(eigenvalue)
    engen_vector = calc_engenvector_2d(variance_coveriance_matrix, eigenvalue)
    print("- 固有ベクトル -")
    print("z1 = " + str(engen_vector[0][0]) + " x1 " + ("+" if engen_vector[0][1] > 0 else "") + str(engen_vector[0][1]) + " x2")
    print("z2 = " + str(engen_vector[1][0]) + " x1 " + ("+" if engen_vector[1][1] > 0 else "") + str(engen_vector[1][1]) + " x2")
結果
> python MultipleRegressionAnalysis_3.py
- 分散共分散行列 -
[[  2.36693333   2.99921333]
 [  2.99921333  10.61296533]]
- 固有値 -
[11.58843357427205, 1.3914650923945988]
- 固有ベクトル -
z1 = 0.30929366776230544 x1 +0.9509665751655736 x2
z2 = 0.9509665751655736 x1 -0.3092936677623057 x2
1
1
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
1
1

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?