April 21, 2023

A Quick Dashboard for Amazon Products

Building a Shiny dashboard over a Kaggle dataset of Amazon India products — categories ranked by ratings and reviews.

Abstract

This project focuses on the development of a Shiny application to analyze Amazon products in India. Using a dataset sourced from Kaggle, the application aims to create an informative dashboard showcasing categories with the highest average ratings and reviews. Despite encountering challenges with product categorization, the project prioritizes efficiency and design improvements.

Introduction

The objective of this project is to create a Shiny application that explores the Amazon products dataset from India. The dataset, obtained from Kaggle, presented challenges during development, particularly regarding product categories. However, the project’s primary focus was on creating a concise and informative dashboard displaying categories with the highest average ratings and reviews.

Compared to previous Shiny app development experiences, this project demonstrated improved efficiency, taking significantly less time to complete. Furthermore, advanced CSS theming was implemented to enhance the overall design of the application. While additional features and visualizations could be incorporated, the decision was made to leave them for future exploration.

Questions

What is the average rating for each category and subcategory?
What is the average review for each category and subcategory?

Further Improvement Questions

Is there a correlation between the number of ratings and the product rating?
What is the average discount percentage for each main category and subcategory?
What is the price range for each main category and subcategory?
Which products have the highest ratings and how do they compare in terms of price and number of ratings?

Conclusion

This project served as a rewarding experience, providing opportunities to enhance data analysis and Shiny app development skills. The Shiny application developed enables users to analyze Amazon products in India and offers insights into categories with the highest average ratings and reviews. Despite encountering challenges related to product categorization, the project prioritized efficiency and introduced improvements in design. Future exploration of additional features and visualizations remains open for further development and enhancement.

Fullscreen

Appendix

Load Libraries

Code

library(shiny)
library(tidyverse)
library(readr)
library(stringr)
library(plotly)
library(DT)
options(scipen = 999)

Part 1: Data Wrangling Multiple Datasets

Code

# ORIGINAL DATA WRANGLING 

datasets <- as.data.frame(list.files(path = "~/Documents/ShinyApps/AmazonProducts/AmazonProductApp", pattern = "csv"))
colnames(datasets) <- "Datasets"

# Combine all the datasets
for (i in length(nrow(datasets))){
 combinedDs <- read_csv(datasets[[i]])
}

amazonProducts <- combinedDs %>%
 mutate(Name = name,
       MainCategory = factor(str_to_title((sort(main_category)))),
       SubCategory = factor(sort(sub_category)),
       ProductImage = image,
       ProductRating = as.numeric(ratings),
       NumberOfRatings = as.numeric(gsub("\\,","",no_of_ratings)),
       DiscountPrice = round(as.numeric(gsub("[\\₹,]", "", discount_price)) / 81.85, 2), # convert from Rupee to USD
       Price = round(as.numeric(gsub("[\\₹,]", "", actual_price)) / 81.85, 2),           # convert from Rupee to USD
       ProductLink = link) %>%
 select(-c(name, 
           main_category, 
           sub_category, 
           image, 
           ratings, 
           no_of_ratings, 
           discount_price, 
           actual_price,
           link)) %>% 
 drop_na() %>%
 filter(!str_detect(SubCategory, "^All "))

# amazonProducts %>%
# write_csv("amazonProducts.csv")

Reload Data

Code

products <- read_csv("/data/amazonproducts.csv")

Part 2: Data Wrangling For Visualization

Code

products <- products %>% mutate(DiscountPrice = round(DiscountPrice82.04,2), Price = round(Price82.04,2))

#For Plot 1
ratings <- products %>% select(-c(ProductImage, Name, ProductLink)) %>% group_by(MainCategory,SubCategory) %>% summarise(AverageRating = mean(ProductRating)) %>% ungroup()

reviews <- products %>% select(-c(ProductImage, Name, ProductLink)) %>% group_by(MainCategory,SubCategory) %>% summarise(AverageReview = mean(NumberOfRatings)) %>% ungroup()

#For Plot 2 
top10Products <- products %>% filter(ProductRating > 4.5, NumberOfRatings > 50) %>% group_by(SubCategory) %>% arrange(desc(ProductRating)) %>% slice(1:10) %>% select(-c(ProductImage, Name, ProductLink))

#unique(products$MainCategory)

Color Theming

Code

#Plot 1 
num_colors <- 21 
colors <- c("#f2f2f2", "#ff9900") 
pal1 <- colorRampPalette(colors)(num_colors)

print(pal1)

#Plot 2 
num_colors <- 21 
colors <- colors <- c("#f2f2f2","#00a8e1") 
pal2 <- colorRampPalette(colors)(num_colors)

Part 1: Shiny UI

Code

# Define UI for application 
ui <- fluidPage(
  
  #Background CSS
  tags$head(tags$style(HTML('
  @import url(https://fonts.googleapis.com/css?family=Montserrat&display=swap);
    body {
      font-family: Montserrat, sans-serif;
      background-color: #FF9900;
    }
    .dataTables_wrapper {
      background-color: #fff;
    }
    .sidebar {
      background-color: #fff;
      width: 3/12;
      height: 2/12;
    }
    .nav-tabs > li > a {
      color: black;
      background-color: #00a8e1;
      border-color: #00a8e1;
    }'))),
  
  # Application title
  titlePanel("Amazon Inc. Product Dashboard (EN.)"),
  
  # Sidebar
  sidebarLayout(
    sidebarPanel(
      selectInput("MainCategoryChoice", 
                  label = h3("Select Category:"), 
                  choices = unique(products$MainCategory), 
                  selected = "Accessories"),
      
      uiOutput("SubCategoryChoice")
    ),
    
    # Tabs
    mainPanel(
      tabsetPanel(
        tabPanel("Plot", 
                 plotlyOutput("RatingsPlot"),
                 plotlyOutput("ReviewsPlot")),
        tabPanel("Data", dataTableOutput("myDataTable"))
      )
    )
  )
)

Part 2: Shiny Server

Code

# Define server logic required to draw Dashboard
server <- function(input, output) {
  
  # Subcategory choices
  output$SubCategoryChoice <- renderUI({
    subcategories <- unique(products$SubCategory[products$MainCategory == input$MainCategoryChoice])
    checkboxGroupInput("SubCategoryChoice", 
                       label = h3("Select Subcategories:"), 
                       choices = subcategories, 
                       selected = subcategories)
  })
  
  # Plot 1: Product Rating
  output$RatingsPlot <- renderPlotly({
    ratings %>%
      filter(MainCategory == input$MainCategoryChoice,
             SubCategory %in% input$SubCategoryChoice) %>% 
      mutate(SubCategory_ordered = factor(SubCategory, 
                                          levels = unique(SubCategory[order(AverageRating)]))) %>%
      plot_ly(x = ~SubCategory_ordered, 
              y = ~round(AverageRating,2),
              type = 'bar',
              marker = list(color = ~pal1[SubCategory_ordered])) %>%
      add_annotations(x = ~SubCategory_ordered,
                      y = ~round(AverageRating,2),
                      text = ~paste0(round(AverageRating,2)),
                      font = list(color = 'black', 
                                  size = 10),
                      showarrow = FALSE,
                      yshift = 5) %>%
      layout(title = paste0("Average Product Rating For ",input$MainCategoryChoice),
             xaxis = list(title = "", tickangle = 45),
             yaxis = list(title = ""),
             showlegend = FALSE,
             margin = list(t = 50,
                           l = 50,
                           r = 50,
                           b = 50)) 
  })
  
  # Plot 2: Product Reviews
  output$ReviewsPlot <- renderPlotly({
    reviews %>%
      filter(MainCategory == input$MainCategoryChoice,
             SubCategory %in% input$SubCategoryChoice) %>% 
      mutate(SubCategory_ordered = factor(SubCategory, 
                                          levels = unique(SubCategory[order(AverageReview)]))) %>%
      plot_ly(x = ~SubCategory_ordered, 
              y = ~round(AverageReview),
              type = 'bar',
              marker = list(color = ~pal2[SubCategory_ordered])) %>%
      add_annotations(x = ~SubCategory_ordered,
                      y = ~round(AverageReview),
                      text = ~paste0(round(AverageReview)),
                      font = list(color = 'black', 
                                  size = 10),
                      showarrow = FALSE,
                      yshift = 5) %>%
      layout(title = paste0("Average Number of Reviews For ",input$MainCategoryChoice),
             xaxis = list(title = "", tickangle = 45),
             yaxis = list(title = ""),
             showlegend = FALSE,
             margin = list(t = 50,
                           l = 50,
                           r =50,
                           b = 50)) 
  })
  
  output$myDataTable <- DT::renderDataTable({
    products %>%
      filter(MainCategory == input$MainCategoryChoice,
             SubCategory %in% input$SubCategoryChoice) %>%
      mutate(ProductImage = sprintf('<img src="%s" width="75px"/>', ProductImage)) %>%
      DT::datatable(., escape = FALSE, options = list(
        pageLength = 10,
        lengthMenu = c(5, 10, 25),
        scrollY = "600px",
        scrollX = TRUE
      )) %>%
      DT::formatStyle(columns = colnames(products), 
                      backgroundColor = styleEqual(c("green", "white"), c("rgb(51, 102, 0)", "rgb(255, 255, 255)")))
  })
  
  
}

# Run the application 
shinyApp(ui = ui, server = server)

References

Chang, Winston, Joe Cheng, JJ Allaire, Carson Sievert, Barret Schloerke, Yihui Xie, Jeff Allen, Jonathan McPherson, Alan Dipert, and Barbara Borges. 2024. Shiny: Web Application Framework for r. https://shiny.posit.co/.

Grolemund, Garrett, and Hadley Wickham. 2011. “Dates and Times Made Easy with lubridate.” Journal of Statistical Software 40 (3): 1–25. https://www.jstatsoft.org/v40/i03/.

Müller, Kirill, and Hadley Wickham. 2023. Tibble: Simple Data Frames. https://tibble.tidyverse.org/.

Sievert, Carson. 2020. Interactive Web-Based Data Visualization with r, Plotly, and Shiny. Chapman; Hall/CRC. https://plotly-r.com.

Sievert, Carson, Chris Parmer, Toby Hocking, Scott Chamberlain, Karthik Ram, Marianne Corvellec, and Pedro Despouy. 2024. Plotly: Create Interactive Web Graphics via Plotly.js. https://plotly-r.com.

Spinu, Vitalie, Garrett Grolemund, and Hadley Wickham. 2024. Lubridate: Make Dealing with Dates a Little Easier. https://lubridate.tidyverse.org.

Wickham, Hadley. 2016. Ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. https://ggplot2.tidyverse.org.

———. 2023a. Forcats: Tools for Working with Categorical Variables (Factors). https://forcats.tidyverse.org/.

———. 2023b. Stringr: Simple, Consistent Wrappers for Common String Operations. https://stringr.tidyverse.org.

———. 2023c. Tidyverse: Easily Install and Load the Tidyverse. https://tidyverse.tidyverse.org.

Wickham, Hadley, Mara Averick, Jennifer Bryan, Winston Chang, Lucy D’Agostino McGowan, Romain François, Garrett Grolemund, et al. 2019. “Welcome to the tidyverse.” Journal of Open Source Software 4 (43): 1686. https://doi.org/10.21105/joss.01686.

Wickham, Hadley, Winston Chang, Lionel Henry, Thomas Lin Pedersen, Kohske Takahashi, Claus Wilke, Kara Woo, Hiroaki Yutani, Dewey Dunnington, and Teun van den Brand. 2024. Ggplot2: Create Elegant Data Visualisations Using the Grammar of Graphics. https://ggplot2.tidyverse.org.

Wickham, Hadley, Romain François, Lionel Henry, Kirill Müller, and Davis Vaughan. 2023. Dplyr: A Grammar of Data Manipulation. https://dplyr.tidyverse.org.

Wickham, Hadley, and Lionel Henry. 2025. Purrr: Functional Programming Tools. https://purrr.tidyverse.org/.

Wickham, Hadley, Jim Hester, and Jennifer Bryan. 2024. Readr: Read Rectangular Text Data. https://readr.tidyverse.org.

Wickham, Hadley, Davis Vaughan, and Maximilian Girlich. 2024. Tidyr: Tidy Messy Data. https://tidyr.tidyverse.org.

Xie, Yihui, Joe Cheng, and Xianying Tan. 2024. DT: A Wrapper of the JavaScript Library DataTables. https://github.com/rstudio/DT.