Amazon Bestsellers Data Analysis

library(tidyverse)
library(ggplot2)
library(gghighlight)

#Importing the csv file to a books dataframe
books <- read.csv("bestsellers with categories.csv")

#Viewing the dataframe
head(books)

#Identifying the column names
names(books)

[1] "Name"        "Author"      "User.Rating" "Reviews"     "Price"      
[6] "Year"        "Genre"      

Correcting Formatting Issues

# Change the column name from User.Rating to User_Rating
names(books)[names(books) == "User.Rating"] <- "User_Rating"

Missing Values

#Identifying total number of missing values
sum(is.na(books))

[1] 0

There are no missing values in the data.

Statistics

#Identifying Number of books
total_books = count(books)

print(paste("There are",total_books,"books in total."))

[1] "There are 550 books in total."

#Identifying Number of books in each genre
count(books,Genre)

#Identifying the average user rating, rounded to 2 decimal places
avg_rating <- round(mean(books$User_Rating),digits =2)

#Identifying the average price, rounded to 2 decimal places
avg_price <- round(mean(books$Price),digits =2)

#Identifying the average number of reviews, rounded to 2 decimal places
avg_reviews <- round(mean(books$Reviews),digits =0)

#Identifying the highest user rating
max_rating <- max(books$User_Rating)

#Identifying the highest price
max_price <- max(books$Price)

#Identifying the highest price
max_reviews<- max(books$Reviews)

cat("For books in the Bestseller list: \nAverage user rating is",avg_rating,"and the highest user rating is",max_rating,"\nAverage price of $",avg_price,"and the highest price is $",max_price, "\nAverage number of reviews is",avg_reviews,"and the highest number of reviews is",max_reviews)

For books in the Bestseller list: 
Average user rating is 4.62 and the highest user rating is 4.9 
Average price of $ 13.1 and the highest price is $ 105 
Average number of reviews is 11953 and the highest number of reviews is 87841

#Grouping by genre
genre <- books %>% group_by(Genre)

#Summarizing average price for for books each year in each genre
genre_stats <- genre %>% summarise(Price = mean(Price))

#Renaming the price column to Average_price
colnames(genre_stats)[which(names(genre_stats) == "Price")] <- "Average_price"

#Summarizing average number of reviews for for books each year in each genre
genre_avg_reviews <- genre %>% summarise(Reviews = mean(Reviews))

#Adding the "Reviews" column from "genre_avg_reviews" to the genre dataframe as "average_reviews" 
genre_stats$Average_number_of_reviews <- genre_avg_reviews$Reviews

#Summarizing average user rating for for books each year in each genre
genre_avg_rating <- genre %>% summarise(User_Rating = mean(User_Rating))

genre_stats$Average_user_rating <- genre_avg_rating$User_Rating

genre_stats

On average,

• Fiction books receive more reviews and higher user ratings than Non-Fiction books.
• Non-Fiction books are more expensive than Fiction books.

Plots

User Rating

#plotting bar graph for user rating vs. number of reviews
ggplot(data=books,aes(x=User_Rating,y=Reviews)) + geom_bar(stat="identity") + labs(title="User Rating vs. Number of Reviews")

• Books that receive a higher user rating recieve more reviews.
• There is a significant increase in the number of reviews for books that have a rating of 4.5 and above.

#Grouping by user rating
by_rating_price <- books %>% group_by(User_Rating)

#Summarizing average price for each user rating value
by_rating_price <- by_rating_price %>% summarise(Price = mean(Price))

#plotting bar graph for user rating vs. price
ggplot(data=by_rating_price,aes(x=User_Rating,y=Price)) + geom_col() + labs(title="User Rating vs. Average Price")

On average, books with a user rating of 4.5 tend to be the most expensive.

Price

#plotting scatter plot for price vs. number of reviews
##geom_point() creates a scatter plot
#facet_wrap(~Genre) groups the data by Genre
#gghighlight(Reviews>25000) will highlight books with more than 25,000 reviews 
ggplot(data=books,aes(x=Price,y=Reviews)) + geom_point() + labs(title="Price vs. Number of Reviews") + gghighlight(Price<=25)

Generally, books priced at or under $25 receive the most reviews.

#plotting scatter plot for price vs. number of reviews
#Highlighting books with more than 25,000 reviews 
ggplot(data=books,aes(x=Price,y=Reviews)) + geom_point() + facet_wrap(~Genre) + labs(title="Price vs. Number of Reviews by Genre") + gghighlight(Reviews>25000)

Highlighted above are the books that received over 25,000 reviews.

Year

#Grouping by year
by_year <- books %>% group_by(Year)

#Summarizing average price for for books each year
by_year_price <- by_year %>% summarise(Price = mean(Price))

#Summarizing average price for for books each year
by_year_reviews <- by_year %>% summarise(Reviews = mean(Reviews))

#plotting bar graph for year vs. average number of reviews
#You have to specify stat="identity" for this kind of dataset.
#labs(title=" ") sets the title
ggplot(data=by_year_reviews,aes(x=Year,y=Reviews)) + geom_col() + labs(title="Average Number of Reviews Per Year")

After mid-2011 the average number of reviews received increased significantly.

#plotting bar graph for year vs. average price
#You have to specify stat="identity" for this kind of dataset.
#labs(title=" ") sets the title
ggplot(data=by_year_price,aes(x=Year,y=Price)) + geom_col() + labs(title="Average Price Per Year")

The average price for an Amazon Top 50 Bestseller has remained between $10 to $15 between 2009-2019.

Summary of Data Analysis

There are 550 books in total.

• Fiction: 240
• Non-Fiction: 310

For books on the Bestseller list in the period from 2009-2019,

• Average user rating is 4.62 and the highest user rating is 4.9
• Average price of $13.10 and the highest price is $105.00
• Average number of reviews is 11953 and the highest number of reviews is 87841
• Fiction books receive more reviews and higher user ratings than Non-Fiction books.
• Non-Fiction books are more expensive than Fiction books.

Relationship between Price, User Rating, Number of Reviews and Year of Release

• The average price for an Amazon Top 50 Bestseller has remained between $10 to $15 between 2009-2019.
• Books priced at or under $25 receive the most reviews, regarless of genre.
• After mid-2011 the average number of reviews received increased significantly.
• Books that have a rating of 4.5 and above receive significantly higher reviews.
• The majority of books receive less than 25,000 reviews.
• Books that receive a higher user rating received more reviews.
• Books with a user rating of 4.5 tend to have the highest average price.