J'ai créé un code pour calculer le coefficient de régression d'une analyse de régression simple, alors veuillez l'utiliser si vous le souhaitez!
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
Column 1 | Column 2 |
---|---|
2.2 | 71 |
4.1 | 81 |
5.5 | 86 |
1.9 | 72 |
3.4 | 77 |
2.6 | 73 |
4.2 | 80 |
3.7 | 81 |
4.9 | 85 |
3.2 | 74 |
dataset = pd.read_csv('test.csv')
Extraire les colonnes avec X comme variable explicative et y comme variable objectif.
X = dataset.iloc[:, :-1].values #Indepand variable
y = dataset.iloc[:, 1].values #Depand variable
# Calculate Mean
Sum_X = sum(X)
N_X = len(X)
Mean_X = Sum_X / N_X
Sum_y = sum(y)
N_y = len(y)
Mean_y = Sum_y / N_y
# Calcuate Deviation
Devi_X = []
for Row_X in X:
Devi_X.append(Row_X - Mean_X)
Devi_y = []
for Row_y in y:
Devi_y.append(Row_y - Mean_y)
# Multiply Deviation X and y
counter = 0
MD_Xy = []
while counter < len(Devi_X):
MD_Xy_Value = Devi_X[counter] * Devi_y[counter]
MD_Xy.append(MD_Xy_Value)
counter += 1
# Sum of Multiply Deviation X and y
SMD_Xy = sum(MD_Xy)
# Squares of Calcuate Deviation
Sq_Devi_X = []
for DX in Devi_X:
Sq_Devi_X.append(DX * DX)
Sq_Devi_y = []
for Dy in Devi_y:
Sq_Devi_y.append(Dy * Dy)
# Sum of Squares of Calcuate Deviation
SSX = sum(Sq_Devi_X)
SSy = sum(Sq_Devi_y)
# Calculate Regression paramator
betaOne = SMD_Xy / SSX
betaZero = Mean_y - betaOne * Mean_X
Recommended Posts