Linear Regression using sklearn

Simple Linear Regression using scikit-learn

Linear Regression is a statistical model used to predict the linear relationship between two or more variables.

Here we are going to demonstrate the linear Regression model using the Scikit-learn library in Python.

Scikit-learn also defined as sklearn is a python library with a lot of efficient tools for machine learning and statistical modeling including classification, regression, clustering, and dimensionality reduction. It features various algorithms like support vector machines, random forests, and k-neighbors.

The dataset used for this model contains the Experience and Salary of Employees. The Salary is based on the Years of Experience of the employee. We are going to derive a linear relationship between the years of experience and the salary.

You can download the dataset here- Dataset

Implementation of the linear regression model using sklearn :-

Step 1: Import all the required libraries for the model.

# importing the required libraries
import pandas as pd

import numpy as np

import matplotlib.pyplot as plt

Step 2: Import the dataset.

# read the dataset using pandas

Data = pd.read_csv(“Salary_Data.csv”)

print(“Data imported successfully”)

The dataset is imported using the read_csv function in the pandas library. If the dataset file is imported successfully it will print “Data imported successfully”

Data.head() # displays the top 5 rows of the data

head() function in NumPy is used to display the number of rows of data we want to display. By default, it will display 5 rows if we didn’t pass any arguments.

Data.info() # Provides information regarding the columns in the data

Step 3: Plotting the data in a graph

Data.plot(x=’YearsExperience’,y=’Salary’,style=’o’)

plt.title(“Salary Prediction with experience”)

plt.xlabel(“Experience in years”)

plt.ylabel(“Salary”)

Plotting the data to know how the values are scattered using the plot function.

output image

# Assigning the data into rows and columns

X = Data.iloc[:, :-1].values

Y = Data.iloc[:, 1].values

Step 4: Splitting the data into training and testing data

# Split the data for train and test

from sklearn.model_selection import train_test_split

x_train,x_test,y_train,y_test=train_test_split(X,Y,test_size=0.2,random_state=0)

To split the data into training and testing sets, import a module train_test_split in the sklearn library.

Step 5: Train the data

#Training the data

from sklearn.linear_model import LinearRegression

regressor = LinearRegression()

regressor.fit(x_train,y_train)

print(“Training sucessful”)

Train the dataset using the fit() method in the LinearRegression function of sklearn. It will print “Training successful” if the dataset is trained by our model.

Step 6: Plot Regression line

# plotting the regression line

line=regressor.coef_*X+regressor.intercept_

plt.scatter(X,Y)

plt.plot(X,line)

plt.show()

# Intecept and coeff of the line

print(‘Intercept of the model:’,regressor.intercept_)

print(‘Coefficient of the line:’,regressor.coef_)

The linear regression model is represented by the equation, y = mx + c
m = coefficient of the line
c = intercept of the line

Therefore the equation for this model is represented as

y = (26780.09)x + 9312.5

print(x_test) #printing the test data

Test data:

Step 7: Predicting the result

y_pred=regressor.predict(x_test)

df=pd.DataFrame({‘actual’:y_test,’predicted’:y_pred})

df

The predict() method is used to predict the result. Here we passed the test data as input.

year = float(input(“Enter number of years : “))

year = np.array(year).reshape(-1, 1)

own_pred=regressor.predict(year)

print(“Predicted Salary = {}”.format(own_pred[0]))

We can also predict the output by giving the input manually.

Step 8: Calculating Error

Mean absolute error (MAE) is a measure of errors between prediction and true values.

	=	prediction values
	=	true value
	=	total number of data

from sklearn import metrics

print(‘Mean Absolute Error:’, metrics.mean_absolute_error(y_test, y_pred))

The mean absolute error is calculated by the mean_absolute_error() method in sklearn.

Mean Absolute Error: 2446.1723690465055

Simple Linear Regression using scikit-learn