Hands-on Exercise 10

Author

Enrico Sebastian

Published

June 25, 2025

Modified

June 27, 2025

31 Information Dashboard Design: R methods

31.1 Overview

By the end of this hands-on exercise, you will be able to:

  • create bullet chart by using ggplot2,

  • create sparklines by using ggplot2 ,

  • build industry standard dashboard by using R Shiny.

31.2 Getting started

For the purpose of this hands-on exercise, the following R packages will be used.

Code 
pacman::p_load(lubridate, ggthemes, reactable,
reactablefmtr, gt, gtExtras, tidyverse)

  • tidyverse provides a collection of functions for performing data science task such as importing, tidying, wrangling data and visualising data. It is not a single package but a collection of modern R packages including but not limited to readr, tidyr, dplyr, ggplot, tibble, stringr, forcats and purrr.

  • lubridate provides functions to work with dates and times more efficiently.

  • ggthemes is an extension of ggplot2. It provides additional themes beyond the basic themes of ggplot2.

  • gtExtras provides some additional helper functions to assist in creating beautiful tables with gt, an R package specially designed for anyone to make wonderful-looking tables using the R programming language.

  • reactable provides functions to create interactive data tables for R, based on the React Table library and made with reactR.

  • reactablefmtr provides various features to streamline and enhance the styling of interactive reactable tables with easy-to-use and highly-customizable functions and themes.

31.3 Importing Microsoft Access database

31.3.1 The data set

For the purpose of this study, a personal database in Microsoft Access mdb format called Coffee Chain will be used.

31.3.2 Importing database into R

In the code chunk below, odbcConnectAccess() of RODBC package is used used to import a database query table into R.

Code 
library(RODBC)
con <- odbcConnectAccess2007('data/Coffee Chain.mdb')
coffeechain <- sqlFetch(con, 'CoffeeChain Query')
write_rds(coffeechain, "data/CoffeeChain.rds")
odbcClose(con)

Note: Before running the code chunk, you need to change the R system to 32bit version. This is because the odbcConnectAccess() is based on 32bit and not 64bit

31.3.3 Data Preparation

The code chunk below is used to import CoffeeChain.rds into R.

Code 
coffeechain <- read_rds("data/rds/CoffeeChain.rds")

Note: This step is optional if coffeechain is already available in R.

The code chunk below is used to aggregate Sales and Budgeted Sales at the Product level.

Code 
product <- coffeechain %>%
  group_by(`Product`) %>%
  summarise(`target` = sum(`Budget Sales`),
            `current` = sum(`Sales`)) %>%
  ungroup()

31.3.4 Bullet chart in ggplot2

The code chunk below is used to plot the bullet charts using ggplot2 functions.

Code 
ggplot(product, aes(Product, current)) + 
  geom_col(aes(Product, max(target) * 1.01),
           fill="grey85", width=0.85) +
  geom_col(aes(Product, target * 0.75),
           fill="grey60", width=0.85) +
  geom_col(aes(Product, target * 0.5),
           fill="grey50", width=0.85) +
  geom_col(aes(Product, current), 
           width=0.35,
           fill = "black") + 
  geom_errorbar(aes(y = target,
                    x = Product, 
                    ymin = target,
                    ymax= target), 
                width = .4,
                colour = "red",
                size = 1) +
  coord_flip()

31.4 Plotting sparklines using ggplot2

In this section, you will learn how to plot sparklines by using ggplot2.

31.4.1 Preparing the data

Code 
sales_report <- coffeechain %>%
  filter(Date >= "2013-01-01") %>%
  mutate(Month = month(Date)) %>%
  group_by(Month, Product) %>%
  summarise(Sales = sum(Sales)) %>%
  ungroup() %>%
  select(Month, Product, Sales)

The code chunk below is used to compute the minimum, maximum and end othe the month sales.

Code 
mins <- group_by(sales_report, Product) %>% 
  slice(which.min(Sales))
maxs <- group_by(sales_report, Product) %>% 
  slice(which.max(Sales))
ends <- group_by(sales_report, Product) %>% 
  filter(Month == max(Month))

The code chunk below is used to compute the 25 and 75 quantiles.

Code 
quarts <- sales_report %>%
  group_by(Product) %>%
  summarise(quart1 = quantile(Sales, 
                              0.25),
            quart2 = quantile(Sales, 
                              0.75)) %>%
  right_join(sales_report)

31.4.2 sparklines in ggplot2

The code chunk used.

Code 
ggplot(sales_report, aes(x=Month, y=Sales)) + 
  facet_grid(Product ~ ., scales = "free_y") + 
  geom_ribbon(data = quarts, aes(ymin = quart1, max = quart2), 
              fill = 'grey90') +
  geom_line(size=0.3) +
  geom_point(data = mins, col = 'red') +
  geom_point(data = maxs, col = 'blue') +
  geom_text(data = mins, aes(label = Sales), vjust = -1) +
  geom_text(data = maxs, aes(label = Sales), vjust = 2.5) +
  geom_text(data = ends, aes(label = Sales), hjust = 0, nudge_x = 0.5) +
  geom_text(data = ends, aes(label = Product), hjust = 0, nudge_x = 1.0) +
  expand_limits(x = max(sales_report$Month) + 
                  (0.25 * (max(sales_report$Month) - min(sales_report$Month)))) +
  scale_x_continuous(breaks = seq(1, 12, 1)) +
  scale_y_continuous(expand = c(0.1, 0)) +
  theme_tufte(base_size = 3, base_family = "Helvetica") +
  theme(axis.title=element_blank(), axis.text.y = element_blank(), 
        axis.ticks = element_blank(), strip.text = element_blank())

31.5 Static Information Dashboard Design: gt and gtExtras methods

In this section, you will learn how to create static information dashboard by using gt and gtExtras packages. Before getting started, it is highly recommended for you to visit the webpage of these two packages and review all the materials provided on the webpages at least once. You done not have to understand and remember everything provided but at least have an overview of the purposes and functions provided by them.

31.5.1 Plotting a simple bullet chart

In this section, you will learn how to prepare a bullet chart report by using functions of gt and gtExtras packages.

Code 
product %>%
  gt::gt() %>%
  gt_plt_bullet(column = current, 
              target = target, 
              width = 60,
              palette = c("lightblue", 
                          "black")) %>%
  gt_theme_538()

31.6 sparklines: gtExtras method

Before we can prepare the sales report by product by using gtExtras functions, code chunk below will be used to prepare the data.

Code 
report <- coffeechain %>%
  mutate(Year = year(Date)) %>%
  filter(Year == "2013") %>%
  mutate (Month = month(Date, 
                        label = TRUE, 
                        abbr = TRUE)) %>%
  group_by(Product, Month) %>%
  summarise(Sales = sum(Sales)) %>%
  ungroup()

It is important to note that one of the requirement of gtExtras functions is that almost exclusively they require you to pass data.frame with list columns. In view of this, code chunk below will be used to convert the report data.frame into list columns.

Code 
report %>%
  group_by(Product) %>%
  summarize('Monthly Sales' = list(Sales), 
            .groups = "drop")

31.6.1 Plotting Coffechain Sales report

Code 
report %>%
  group_by(Product) %>%
  summarize('Monthly Sales' = list(Sales), 
            .groups = "drop") %>%
   gt() %>%
   gt_plt_sparkline('Monthly Sales',
                    same_limit = FALSE)

31.6.2 Adding statistics

First, calculate summary statistics by using the code chunk below.

Code 
report %>% 
  group_by(Product) %>% 
  summarise("Min" = min(Sales, na.rm = T),
            "Max" = max(Sales, na.rm = T),
            "Average" = mean(Sales, na.rm = T)
            ) %>%
  gt() %>%
  fmt_number(columns = 4,
    decimals = 2)

31.6.3 Combining the data.frame

Next, use the code chunk below to add the statistics on the table.

Code 
spark <- report %>%
  group_by(Product) %>%
  summarize('Monthly Sales' = list(Sales), 
            .groups = "drop")
Code 
sales <- report %>% 
  group_by(Product) %>% 
  summarise("Min" = min(Sales, na.rm = T),
            "Max" = max(Sales, na.rm = T),
            "Average" = mean(Sales, na.rm = T)
            )
Code 
sales_data = left_join(sales, spark)

31.6.4 Plotting the updated data.table

Code 
sales_data %>%
  gt() %>%
  gt_plt_sparkline('Monthly Sales',
                   same_limit = FALSE)

31.6.5 Combining bullet chart and sparklines

Similarly, we can combining the bullet chart and sparklines using the steps below.

Code 
bullet <- coffeechain %>%
  filter(Date >= "2013-01-01") %>%
  group_by(`Product`) %>%
  summarise(`Target` = sum(`Budget Sales`),
            `Actual` = sum(`Sales`)) %>%
  ungroup()
Code 
sales_data = sales_data %>%
  left_join(bullet)
Code 
sales_data %>%
  gt() %>%
  gt_plt_sparkline('Monthly Sales') %>%
  gt_plt_bullet(column = Actual, 
                target = Target, 
                width = 28,
                palette = c("lightblue", 
                          "black")) %>%
  gt_theme_538()

31.7 Interactive Information Dashboard Design: reactable and reactablefmtr methods

In this section, you will learn how to create interactive information dashboard by using reactable and reactablefmtr packages. Before getting started, it is highly recommended for you to visit the webpage of these two packages and review all the materials provided on the webpages at least once. You done not have to understand and remember everything provided but at least have an overview of the purposes and functions provided by them.

In order to build an interactive sparklines, we need to install dataui R package by using the code chunk below.

Code 
remotes::install_github("timelyportfolio/dataui")

Next, you all need to load the package onto R environment by using the code chunk below.

Code 
library(dataui)

31.7.1 Plotting interactive sparklines

Similar to gtExtras, to plot an interactive sparklines by using reactablefmtr package we need to prepare the list field by using the code chunk below.

Code 
report <- report %>%
  group_by(Product) %>%
  summarize(`Monthly Sales` = list(Sales))

Next, react_sparkline will be to plot the sparklines as shown below.

Code 
reactable(
  report,
  columns = list(
    Product = colDef(maxWidth = 200),
    `Monthly Sales` = colDef(
      cell = react_sparkline(report)
    )
  )
)

31.7.2 Changing the pagesize

By default the pagesize is 10. In the code chunk below, arguments defaultPageSize is used to change the default setting.

Code 
reactable(
  report,
  defaultPageSize = 13,
  columns = list(
    Product = colDef(maxWidth = 200),
    `Monthly Sales` = colDef(
      cell = react_sparkline(report)
    )
  )
)

31.7.3 Adding points and labels

In the code chunk below highlight_points argument is used to show the minimum and maximum values points and label argument is used to label first and last values.

Code 
reactable(
  report,
  defaultPageSize = 13,
  columns = list(
    Product = colDef(maxWidth = 200),
    `Monthly Sales` = colDef(
      cell = react_sparkline(
        report,
        highlight_points = highlight_points(
          min = "red", max = "blue"),
        labels = c("first", "last")
        )
    )
  )
)

31.7.4 Adding reference line

In the code chunk below statline argument is used to show the mean line.

Code 
reactable(
  report,
  defaultPageSize = 13,
  columns = list(
    Product = colDef(maxWidth = 200),
    `Monthly Sales` = colDef(
      cell = react_sparkline(
        report,
        highlight_points = highlight_points(
          min = "red", max = "blue"),
        statline = "mean"
        )
    )
  )
)

31.7.5 Adding bandline

Instead adding reference line, bandline can be added by using the bandline argument.

Code 
reactable(
  report,
  defaultPageSize = 13,
  columns = list(
    Product = colDef(maxWidth = 200),
    `Monthly Sales` = colDef(
      cell = react_sparkline(
        report,
        highlight_points = highlight_points(
          min = "red", max = "blue"),
        line_width = 1,
        bandline = "innerquartiles",
        bandline_color = "green"
        )
    )
  )
)

31.7.6 Changing from sparkline to sparkbar

Instead of displaying the values as sparklines, we can display them as sparkbars as shiwn below.

Code 
reactable(
  report,
  defaultPageSize = 13,
  columns = list(
    Product = colDef(maxWidth = 200),
    `Monthly Sales` = colDef(
      cell = react_sparkbar(
        report,
        highlight_bars = highlight_bars(
          min = "red", max = "blue"),
        bandline = "innerquartiles",
        statline = "mean")
    )
  )
)

31.8 Reference