heatmap Mark

library(mosaicr)

The heatmap mark creates rectangular heatmaps where color intensity represents data values. It’s perfect for visualizing matrices, correlation data, or any grid-based quantitative information.

Basic Example

# Create a simple heatmap dataset
heatmap_data <- expand.grid(
  x = 1:8,
  y = 1:6
)
heatmap_data$value <- sin(heatmap_data$x * 0.5) * cos(heatmap_data$y * 0.3) * 10

spec <- list(
  plot = list(
    list(
      mark = "rect",
      data = list(from = "heatmap_data"),
      x = "x",
      y = "y",
      fill = list(column = "value")
    )
  )
)

mosaic(spec, heatmap_data = heatmap_data)

Customization Options

Correlation Matrix

# Create a correlation matrix visualization
set.seed(123)
vars <- c("Var1", "Var2", "Var3", "Var4", "Var5")
n_vars <- length(vars)

# Create correlation matrix data
correlation_data <- expand.grid(
  x = vars,
  y = vars
)

# Generate realistic correlation values
correlation_data$correlation <- c(
  1.0, 0.3, -0.2, 0.7, 0.1,    # Var1 correlations
  0.3, 1.0, 0.5, -0.1, 0.8,    # Var2 correlations
  -0.2, 0.5, 1.0, 0.2, 0.4,    # Var3 correlations
  0.7, -0.1, 0.2, 1.0, -0.3,   # Var4 correlations
  0.1, 0.8, 0.4, -0.3, 1.0     # Var5 correlations
)

spec_corr <- list(
  plot = list(
    list(
      mark = "rect",
      data = list(from = "correlation_data"),
      x = "x",
      y = "y",
      fill = "correlation",
      stroke = "white",
      strokeWidth = 1
    )
  ),
  colorScheme = "rdbu",
  xLabel = "Variables",
  yLabel = "Variables"
)

mosaic(spec_corr, correlation_data = correlation_data)

Time Series Heatmap

# Create a time-based heatmap (e.g., hourly data by day)
days <- c("Mon", "Tue", "Wed", "Thu", "Fri", "Sat", "Sun")
hours <- 0:23

time_data <- expand.grid(
  day = days,
  hour = hours
)

# Simulate activity patterns (higher during day, weekends different)
set.seed(456)
time_data$activity <- with(time_data, {
  base_activity <- ifelse(hour >= 6 & hour <= 22,
                         sin((hour - 6) * pi / 16) * 50 + 30,
                         runif(length(hour), 5, 15))

  # Weekend modifier
  weekend_modifier <- ifelse(day %in% c("Sat", "Sun"), 1.2, 1.0)

  pmax(0, base_activity * weekend_modifier + rnorm(length(hour), 0, 5))
})

spec_time <- list(
  plot = list(
    list(
      mark = "rect",
      data = list(from = "time_data"),
      x = "hour",
      y = "day",
      fill = list(column = "activity"),
      stroke = "white",
      strokeWidth = 0.5
    )
  ),
  colorScheme = "plasma",
  xLabel = "Hour of Day",
  yLabel = "Day of Week"
)

mosaic(spec_time, time_data = time_data)

Geographic Grid

# Simulate a geographic heatmap
grid_size <- 10
geo_data <- expand.grid(
  longitude = seq(-5, 5, length.out = grid_size),
  latitude = seq(-5, 5, length.out = grid_size)
)

# Create a pattern that resembles geographic data
geo_data$temperature <- with(geo_data, {
  # Distance from center
  distance <- sqrt(longitude^2 + latitude^2)
  # Temperature decreases with distance, with some noise
  25 - 2 * distance + rnorm(nrow(geo_data), 0, 1)
})

spec_geo <- list(
  plot = list(
    list(
      mark = "rect",
      data = list(from = "geo_data"),
      x = "longitude",
      y = "latitude",
      fill = list(column = "temperature"),
      stroke = "none"
    )
  ),
  colorScheme = "turbo",
  xLabel = "Longitude",
  yLabel = "Latitude"
)

mosaic(spec_geo, geo_data = geo_data)

Annotated Heatmap

# Small heatmap with text annotations
small_data <- expand.grid(
  category_x = c("A", "B", "C", "D"),
  category_y = c("Type 1", "Type 2", "Type 3")
)

small_data$value <- c(23, 45, 12, 67, 34, 56, 78, 29, 41, 33, 52, 18)

spec_annotated <- list(
  plot = list(
    list(
      mark = "rect",
      data = list(from = "small_data"),
      x = "category_x",
      y = "category_y",
      fill = list(column = "value"),
      stroke = "white",
      strokeWidth = 2
    ),
    list(
      mark = "text",
      data = list(from = "small_data"),
      x = "category_x",
      y = "category_y",
      text = "value",
      fill = "white",
      fontSize = 12
    )
  ),
  colorScheme = "viridis"
)

mosaic(spec_annotated, small_data = small_data)

Use Cases

• Correlation Analysis: Visualizing correlation matrices between variables
• Time Patterns: Showing patterns across time periods (hourly, daily, seasonal)
• Geographic Data: Displaying spatial patterns and regional differences
• Performance Metrics: Showing performance across different categories or time periods
• Scientific Data: Visualizing any matrix-structured quantitative data