Question

Python:

Create a class defined for Regression. Class attributes are data points for x, y, the slope and the intercept for the regression line. Define an instance method to find the regression line parameters (slope and intercept). Plot all data points on the graph. Plot the regression line on the same plot.

Answer 1

import numpy as np
import matplotlib.pyplot as plt
class Regression:
   #constructor to initialize the values
   def __init__(self, x, y):
       self.x = x
       self.y = y
   def calculateParameters(self):
       #calculate the means
       x_mean = np.mean(x)
       y_mean = np.mean(y)
       #calculate the size of dataset
       n = np.size(x)
       #calculate the slope of dataset
       self.slope = (np.sum(y*x) - (n*y_mean*x_mean))/(np.sum(x*x) - (n*x_mean*x_mean))
       #calculate the intercept
       self.intercept = y_mean - (self.slope * x_mean)
   def plot_data(self):
       #plotting the data and regression line
       #plot the data points
       plt.scatter(x, y)
       #calculate the points on line
       line = [(self.slope * i) + self.intercept for i in x]
       #plot the line
       plt.plot(x, line)
       #give the lables
       plt.xlabel("x values")
       plt.ylabel("y values")
       plt.title("Regression line")
       plt.show()
#consider two numpy arrays
#this will be considered as dataset
#numpy array for x
x = np.array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
#numpy array for x
y = np.array([1, 3, 2, 5, 7, 8, 8, 9, 10, 12])
#create object for regression
reg = Regression(x, y)
#calculate the slope and intercept
reg.calculateParameters()
#print the values of parameters
print(reg.slope)
print(reg.intercept)
#plot the data
reg.plot_data()  
      
      

The dataset can be loaded from any csv file or can create simple data set to test it.