Bar, Pie and Word Cloud Examples • maidr

Bar and pie charts are the categorical family: one navigable term per bar, wedge or word, moved through with the Left and Right arrow keys, with the value of each announced and mapped to pitch when sonification is on. Bar and pie charts are stable plot types in both ggplot2 and Base R (see “Supported plot types” in the README); the word cloud is an experimental Base R type whose reading may change without a deprecation period. The examples hub lists every other plot family.

Bar Plots

Simple Bar Chart

A simple bar chart compares values across categories. Each bar represents one category with its height proportional to its value.

ggplot2

bar_data <- data.frame(
  Product = c("Laptop", "Tablet", "Phone", "Monitor"),
  Sales = c(150, 230, 180, 290)
)

p <- ggplot(bar_data, aes(x = Product, y = Sales)) +
  geom_bar(stat = "identity", fill = "steelblue") +
  labs(
    title = "Product Sales by Category",
    x = "Product",
    y = "Sales (units)"
  ) +
  theme_minimal()

p

Base R

products <- c("Laptop", "Tablet", "Phone", "Monitor")
sales <- c(150, 230, 180, 290)

barplot(sales,
  names.arg = products,
  col = "steelblue",
  main = "Product Sales by Category",
  xlab = "Product",
  ylab = "Sales (units)"
)

Dodged / Grouped Bar Chart

A dodged bar chart places bars for each sub-group side by side, making it easy to compare values within and across categories.

ggplot2

dodged_data <- data.frame(
  Region = rep(c("North", "South", "East"), each = 2),
  Quarter = rep(c("Q1", "Q2"), 3),
  Revenue = c(120, 150, 200, 180, 160, 210)
)

p <- ggplot(dodged_data, aes(x = Region, y = Revenue, fill = Quarter)) +
  geom_bar(stat = "identity", position = position_dodge(width = 0.8)) +
  labs(title = "Quarterly Revenue by Region") +
  scale_fill_manual(values = c("steelblue", "coral")) +
  theme_minimal()

p

Base R

revenue_matrix <- matrix(c(120, 150, 200, 180, 160, 210), nrow = 2)
rownames(revenue_matrix) <- c("Q1", "Q2")
barplot(revenue_matrix,
  beside = TRUE,
  names.arg = c("North", "South", "East"),
  col = c("steelblue", "coral"),
  legend.text = rownames(revenue_matrix),
  main = "Quarterly Revenue by Region",
  xlab = "Region",
  ylab = "Revenue"
)

Stacked Bar Chart

A stacked bar chart layers sub-groups on top of each other within each category, showing both individual contributions and the total.

ggplot2

stacked_data <- data.frame(
  Year = rep(c("2022", "2023", "2024"), each = 3),
  Source = rep(c("Solar", "Wind", "Hydro"), 3),
  Output = c(40, 30, 50, 55, 45, 48, 70, 60, 52)
)

p <- ggplot(stacked_data, aes(x = Year, y = Output, fill = Source)) +
  geom_bar(stat = "identity", position = position_stack()) +
  labs(
    title = "Renewable Energy Output by Source",
    y = "Output (GWh)"
  ) +
  scale_fill_manual(values = c("#2ecc71", "#3498db", "#9b59b6")) +
  theme_minimal()

p

Base R

energy_matrix <- matrix(
  c(40, 30, 50, 55, 45, 48, 70, 60, 52),
  nrow = 3
)
rownames(energy_matrix) <- c("Solar", "Wind", "Hydro")
barplot(energy_matrix,
  beside = FALSE,
  names.arg = c("2022", "2023", "2024"),
  col = c("#2ecc71", "#3498db", "#9b59b6"),
  legend.text = rownames(energy_matrix),
  main = "Renewable Energy Output by Source",
  xlab = "Year",
  ylab = "Output (GWh)"
)

Pie Chart

A pie chart splits a whole into wedges, one per category, each wedge’s angle proportional to its share. maidr announces every slice by name and value, and derives its percentage of the total for you.

ggplot2

ggplot2 has no pie geom: a pie is a single stacked column bent around by coord_polar("y"), so x is the literal "" and the categories go on fill. coord_radial(theta = "y") works the same way. Note that coord_polar("x") draws a coxcomb instead, which maidr still describes as a bar chart.

market_share <- data.frame(
  Browser = c("Chrome", "Safari", "Edge", "Firefox"),
  Share = c(64, 19, 5, 3)
)

p <- ggplot(market_share, aes(x = "", y = Share, fill = Browser)) +
  geom_col(width = 1) +
  coord_polar("y") +
  labs(
    title = "Browser Market Share",
    fill = "Browser",
    y = "Share (%)"
  ) +
  theme_void()

p

Base R

shares <- c(Chrome = 64, Safari = 19, Edge = 5, Firefox = 3)
pie(shares,
  col = c("#3498db", "#2ecc71", "#9b59b6", "#e67e22"),
  main = "Browser Market Share",
  xlab = "Browser",
  ylab = "Share (%)"
)

Word Cloud

A word cloud is the extreme case of a chart that carries real data while being readable only by eye: each term’s weight is drawn as glyph size and written down nowhere on the page. Structurally it is a categorical label and a magnitude, so maidr reads it as a term and its number.

Experimental. word_cloud is one of the experimental layer types. It has not been through a user study, and it may change without a deprecation period. See the experimental table in the README.

Note: Word clouds are a Base R feature. ggplot2 has no word cloud geom, and the packages that add one draw through their own devices rather than through a layer maidr can read.

Requires the {wordcloud} package. Install with install.packages("wordcloud").

Base R

The counts you pass survive into the reading, so the weight axis honestly says Occurrences. (The Python binding cannot: wordcloud.WordCloud divides every frequency by the largest and keeps only the ratio, so py-maidr announces a relative frequency. Same chart, two different honest readings.)

mentions <- c(
  accessibility = 412, sonification = 300, braille = 250,
  screenreader = 190, keyboard = 155, contrast = 120
)

maidr::wordcloud(
  words = names(mentions),
  freq = mentions,
  min.freq = 1,
  random.order = FALSE,
  colors = c(
    "#3498db", "#2ecc71", "#9b59b6",
    "#e67e22", "#e74c3c", "#16a085"
  )
)

Attach order matters. library(wordcloud) after library(maidr) puts package:wordcloud ahead of package:maidr on the search path, so a bare wordcloud() call reaches the wordcloud package directly and maidr never records it — the chart draws, but show() and save_html() then report that no Base R plot was found. Measured on a three-term cloud:

How it is called Recorded?

library(maidr) then library(wordcloud), bare wordcloud() no

library(wordcloud) then library(maidr), bare wordcloud() yes

maidr::wordcloud() yes

Either attach wordcloud before maidr, or call maidr::wordcloud() explicitly as above, which works in either order.

How it is called	Recorded?
`library(maidr)` then `library(wordcloud)`, bare `wordcloud()`	no
`library(wordcloud)` then `library(maidr)`, bare `wordcloud()`	yes
`maidr::wordcloud()`	yes

No highlighting. wordcloud() draws each term with a bare text() call at a rotation chosen by rot.per, and nothing names those — the exported SVG carries no id attributes at all. So a word cloud is read without a visual highlight, rather than pairing the terms with whatever else happened to resolve.

Two of wordcloud()’s arguments decide which terms are drawn, and the reading replicates both so it never announces a term the chart left out: min.freq (default 3) drops anything rarer, and max.words keeps only the heaviest. wordcloud() also lowers min.freq to 0 when it exceeds every frequency, which is what stops a cloud of rare terms coming out empty; that rule is copied rather than approximated.