How to Excel linear regression of the data matrix operations

Code section:

# -*- coding: utf-8 -*-

import numpy as np

import xlrd

 

path="C:/Users/Desktop/temp/aprotest/2creatmodel/123.xlsx"

data = xlrd.open_workbook(path)

table = data.sheets()[2]

 

cols = table.col_values(0)

cols1 = np.matrix (cols) # list convert the matrix

cols2=np.transpose(cols1)#转置

print cols2

 

cols11 = table.col_values(1)

cols21 = table.col_values(2)

cols31 = table.col_values(3)

cols4 = np.vstack ((cols11, cols21, cols31)) # vstack combined sideways after row, the need transpose

cols5=np.matrix(cols4)

cols6 = np.transpose (cols5) # transpose

print cols6

 

z = np.linalg.inv (cols6.T * cols6) * cols6.T * cols2 # post code duplication show abnormal multiplication

print the

 

Ideas:

First data reading xlrd excel; then convert the data matrix matrix, combined and used for data numpy matrix transpose period; finally do numpy matrix calculation, a linear regression to obtain results.

 

Knowledge points:

The matrix transpose numpy

cols6=np.transpose(cols5)

numpy of data consolidation

cols4 np.vstack = ((cols11, cols21, cols31))

If (cols11, cols21, cols31) has first been used as the matrix transpose, and then merge, it should hstack, rather than vstack.

Can be simply understood as: vstack is the line merger, hstack is the column merger. Specific reference:

https://blog.csdn.net/m0_37602827/article/details/90546621 section 1.2