Scribd
Upload
43 views2 pages

Regression

Uploaded by

Anand Kumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
43 views2 pages

Regression

Uploaded by

Anand Kumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

# Install and load the necessary package 'datarium' for the marketing dataset

install.packages("datarium") # Install the datarium package


library(datarium) # Load the datarium package

# Load and inspect the marketing dataset


data(package="datarium") # List datasets in the datarium package
data("marketing") # Load the marketing dataset
?marketing # Get more information about the dataset
View(marketing) # View the dataset in a tabular format
str(marketing) # Check the structure of the dataset (variables and data types)
summary(marketing) # Get summary statistics (mean, min, max, etc.) for the dataset
plot(marketing, col="red") # Plot the dataset with red points to inspect relationships

# Load the ggplot2 library for advanced visualizations


library(ggplot2)

# Install and load caTools to split the dataset into training and test sets
install.packages("caTools") # Install the caTools package for data splitting
library(caTools) # Load the caTools package

# Split the dataset into training and testing subsets (80% for training, 20% for testing)
set.seed(101) # Set a random seed to ensure reproducibility of the random split
sample = sample.split(1:nrow(marketing), SplitRatio = 0.8) # Create a sample split (80%
training, 20% testing)
sample # Print the sample split to inspect the selection

# Create the training and testing datasets based on the sample split
train = subset(marketing, sample == TRUE) # Training set (80% of data)
test = subset(marketing, sample == FALSE) # Testing set (20% of data)

# Check the dimensions of the training and testing sets


dim(train) # Check the number of rows and columns in the training set
dim(test) # Check the number of rows and columns in the testing set

# Build and evaluate a Simple Linear Regression (LR) model: Predict 'sales' based on
'youtube' spend
model1 <- lm(sales ~ youtube, data=train) # Fit a simple linear regression model using
'youtube' as predictor
summary(model1) # Display a summary of the linear regression model
(coefficients, R-squared, etc.)
pred1 <- predict(model1, test) # Predict sales using the trained model on the test
data

# Build and evaluate a Multiple Linear Regression (MLR) model: Predict 'sales' based on
'youtube', 'facebook', and 'newspaper'
model <- lm(sales ~ youtube + facebook + newspaper, data=train) # Fit a multiple linear
regression model using multiple predictors
summary(model) # Display a summary of the multiple linear
regression model
pred <- predict(model, test) # Predict sales using the multiple regression
model on the test data

# Prepare a new variable for plotting (x-axis as the sequence of test rows)
x_axis <- seq(nrow(test)) # Create an x-axis variable representing row numbers of the test
data

# Combine the predictions and actual sales values into a data frame for easy comparison
df <- data.frame(x_axis, pred, test$sales) # Combine x-axis, predicted sales, and actual
sales into a data frame
View(df) # View the resulting data frame with predictions and actual values

# Create a line plot to compare the predicted sales vs. the actual sales using ggplot2
g <- ggplot(df, aes(x=x_axis)) +
geom_line(aes(y=pred, colour="Predicted")) + # Plot the predicted sales as a line
geom_point(aes(x=x_axis, y=pred, colour="Predicted")) + # Plot the predicted sales as
points
geom_line(aes(y=test$sales, colour="Actual")) + # Plot the actual sales as a line
geom_point(aes(x=x_axis, y=test$sales, colour="Actual")) + # Plot the actual sales as
points
scale_colour_manual("", values = c(Predicted="red", Actual="blue")) # Customize the
colors of the lines and points

You might also like