7  Advanced Data Visualization

7.1 Introduction

Building on the visualization techniques covered in Chapter 6, this chapter explores advanced data visualization methods that can help you communicate complex ecological data more effectively. We’ll focus on creating publication-quality graphics, interactive visualizations, and specialized plots for ecological data.

7.2 Creating Publication-Quality Graphics

7.2.1 Customizing ggplot2 Themes

The ggplot2 package allows extensive customization of plot appearance:

# Load necessary packages
library(ggplot2)
library(dplyr)

# Create a basic scatter plot
base_plot <- ggplot(iris, aes(x = Sepal.Length, y = Petal.Length, color = Species)) +
  geom_point(size = 3, alpha = 0.8) +
  labs(
    title = "Relationship Between Sepal and Petal Length",
    subtitle = "Comparison across three Iris species",
    x = "Sepal Length (cm)",
    y = "Petal Length (cm)",
    caption = "Data source: Anderson's Iris dataset"
  )

# Create a custom theme
custom_theme <- theme_minimal() +
  theme(
    # Text elements
    plot.title = element_text(face = "bold", size = 16),
    plot.subtitle = element_text(size = 12, color = "gray40"),
    axis.title = element_text(face = "bold"),

    # Legend position
    legend.position = "bottom",

    # Grid lines and panel
    panel.grid.minor = element_blank(),
    panel.border = element_rect(color = "gray80", fill = NA, size = 0.5)
  )

# Apply the custom theme
base_plot + custom_theme

Code Explanation

This code demonstrates advanced visualization customization:

1. Base Plot Creation:
   • Uses ggplot() for the foundation
   • Maps variables to aesthetics
   • Adds points with transparency
   • Includes comprehensive labels
2. Theme Customization:
   • Creates a custom theme object
   • Modifies text elements
   • Adjusts legend position
   • Customizes grid and panel appearance
3. Visual Elements:
   • Point size and transparency
   • Color coding by species
   • Axis labels and titles
   • Caption with data source

Results Interpretation

The visualization reveals several key insights:

1. Species Differentiation:
   • Clear clustering of species
   • Distinct morphological patterns
   • Overlap between some species
2. Morphological Relationships:
   • Positive correlation between sepal and petal length
   • Different scaling relationships by species
   • Species-specific size ranges
3. Visual Effectiveness:
   • Clear species separation
   • Readable labels and legend
   • Professional appearance

PROFESSIONAL TIP: Advanced Visualization Techniques

When creating publication-quality visualizations:

1. Theme Design:
   • Create consistent themes
   • Use appropriate font sizes
   • Balance white space
   • Consider journal requirements
2. Aesthetic Choices:
   • Select meaningful colors
   • Use appropriate point sizes
   • Consider transparency
   • Balance visual elements
3. Labeling Strategy:
   • Include clear titles
   • Use informative subtitles
   • Provide data sources
   • Consider audience needs

In ecological research, you might create different custom themes for field guides, scientific publications, presentations, interactive dashboards, or reports for non-scientific audiences, each with design elements optimized for their specific purpose and audience.

7.2.2 Color Palettes for Ecological Data

Choosing appropriate color palettes is crucial for effective visualization:

# Load packages for color palettes
library(RColorBrewer)
library(viridis)

# Display color palettes suitable for ecological data
par(mfrow = c(4, 1), mar = c(2, 6, 2, 1))
display.brewer.pal(8, "YlGn")
display.brewer.pal(8, "BrBG")
display.brewer.pal(11, "RdYlBu")
scales::show_col(viridis(8))

# Apply different color palettes to our plot
plot1 <- base_plot +
  scale_color_brewer(palette = "Set1") +
  custom_theme +
  ggtitle("Color Brewer 'Set1' Palette")

plot2 <- base_plot +
  scale_color_viridis_d() +
  custom_theme +
  ggtitle("Viridis Discrete Palette")

# Display the plots
plot1

plot2

# Create a plot with a sequential color palette for a continuous variable
ggplot(iris, aes(x = Sepal.Length, y = Petal.Length, color = Petal.Width)) +
  geom_point(size = 3, alpha = 0.8) +
  scale_color_viridis_c() +
  custom_theme +
  labs(title = "Iris Dataset with Continuous Color Scale",
       subtitle = "Petal Width mapped to color",
       x = "Sepal Length (cm)",
       y = "Petal Length (cm)",
       color = "Petal Width (cm)")

Code Explanation

This code demonstrates how to use and apply different color palettes for ecological data visualization. Let’s break down the key components:

1. Loading Color Palette Packages
   • We load RColorBrewer, which provides a set of carefully designed color palettes.
   • We also load viridis, which offers perceptually uniform and colorblind-friendly palettes.
2. Displaying Color Palettes
   • We use par(mfrow = c(4, 1)) to set up a 4-row panel for displaying multiple palettes.
   • We display three ColorBrewer palettes that are particularly useful for ecological data:
     • “YlGn” (Yellow-Green): A sequential palette good for representing intensity (e.g., vegetation density)
     • “BrBG” (Brown-Blue-Green): A diverging palette useful for showing deviations from a midpoint (e.g., temperature anomalies)
     • “RdYlBu” (Red-Yellow-Blue): Another diverging palette good for environmental gradients
   • We also display the Viridis palette, which is designed to be perceptually uniform and colorblind-friendly.
3. Applying Palettes to Plots
   • We create two versions of our scatter plot with different color schemes:
     • One using ColorBrewer’s “Set1” palette with scale_color_brewer()
     • Another using the Viridis discrete palette with scale_color_viridis_d()
   • We add our custom theme and appropriate titles to each plot.
4. Continuous Color Mapping
   • We create a third plot that maps a continuous variable (Petal Width) to color.
   • We use scale_color_viridis_c() for a continuous color scale that is perceptually uniform.
   • This demonstrates how to use color to represent a third dimension in your data.

Results Interpretation

Choosing appropriate color palettes is crucial for effective ecological data visualization:

1. Types of Color Palettes:
   • Sequential palettes (like YlGn) are ideal for representing ordered data where values progress from low to high (e.g., species abundance, elevation).
   • Diverging palettes (like BrBG, RdYlBu) are best for data with a meaningful midpoint (e.g., temperature anomalies, pH deviations from neutral).
   • Qualitative palettes (like Set1) are designed for categorical data with no inherent order (e.g., species, habitat types).
2. Colorblind Accessibility:
   • The Viridis palettes are specifically designed to be perceptually uniform and accessible to people with color vision deficiencies.
   • This is particularly important in scientific publications where accurate interpretation of colors is crucial.
3. Ecological Applications:
   • Habitat mapping: Sequential green palettes for vegetation density
   • Climate data: Diverging palettes for temperature or precipitation anomalies
   • Species distribution: Qualitative palettes for different species
   • Environmental gradients: Sequential or diverging palettes for pH, elevation, or pollution levels
4. Practical Considerations:
   • Consider how your visualization will be used (digital display, print, presentation)
   • Test your visualizations with colorblind simulation tools
   • Ensure sufficient contrast between categories for clear differentiation
   • Use complementary visual cues (shapes, sizes) alongside color when possible

By thoughtfully selecting color palettes, you can enhance the interpretability of your ecological visualizations while ensuring they remain accessible to all audiences.

7.2.3 Arranging Multiple Plots

Combining multiple plots can help compare different aspects of your data:

library(patchwork)

# Create individual plots
p1 <- ggplot(iris, aes(x = Species, y = Sepal.Length, fill = Species)) +
  geom_boxplot() +
  labs(title = "Sepal Length by Species",
       x = NULL,
       y = "Sepal Length (cm)") +
  theme_minimal() +
  theme(legend.position = "none")

p2 <- ggplot(iris, aes(x = Species, y = Petal.Length, fill = Species)) +
  geom_boxplot() +
  labs(title = "Petal Length by Species",
       x = NULL,
       y = "Petal Length (cm)") +
  theme_minimal() +
  theme(legend.position = "none")

p3 <- ggplot(iris, aes(x = Sepal.Length, fill = Species)) +
  geom_density(alpha = 0.7) +
  labs(title = "Sepal Length Distribution",
       x = "Sepal Length (cm)",
       y = "Density") +
  theme_minimal()

p4 <- ggplot(iris, aes(x = Petal.Length, fill = Species)) +
  geom_density(alpha = 0.7) +
  labs(title = "Petal Length Distribution",
       x = "Petal Length (cm)",
       y = "Density") +
  theme_minimal()

# Arrange the plots
(p1 + p2) / (p3 + p4) +
  plot_annotation(
    title = "Iris Morphology by Species",
    caption = "Source: Anderson's Iris dataset"
  )

7.3 Interactive Visualizations

Interactive visualizations allow users to explore data in more depth than static plots. In R, packages like plotly and leaflet make it easy to create interactive graphics.

7.3.1 Interactive Plots with plotly

The plotly package allows you to convert ggplot2 graphics to interactive versions:

library(plotly)

# Create a ggplot object
p <- ggplot(iris, aes(x = Sepal.Length, y = Petal.Length, color = Species)) +
  geom_point(size = 3, alpha = 0.7) +
  labs(title = "Relationship between Sepal Length and Petal Length",
       x = "Sepal Length (cm)",
       y = "Petal Length (cm)") +
  theme_minimal() +
  scale_color_viridis_d()

# Check if we're in HTML output mode
if (knitr::is_html_output()) {
  # For HTML output, use the interactive plotly version
  ggplotly(p)
} else {
  # For PDF output, use the static ggplot version
  p + annotate("text", x = 6, y = 6,
               label = "Note: Interactive version available in HTML output",
               fontface = "italic", size = 3)
}

Interactive scatter plot of iris dataset. In the HTML version, you can hover over points to see details, zoom, and pan.

Code Explanation

This code demonstrates how to create an interactive scatter plot using the plotly package. Let’s break down the key components:

1. Creating the Base ggplot2 Plot
   • We start by creating a standard ggplot2 scatter plot of the iris dataset, mapping Sepal Length to the x-axis, Petal Length to the y-axis, and Species to the color aesthetic.
   • We add styling elements like point size, transparency, informative labels, and a clean theme.
   • We use the Viridis color palette for better accessibility.
2. Converting to an Interactive Plot
   • We use the ggplotly() function to convert our ggplot2 object into an interactive plotly visualization.
   • This simple conversion adds several interactive features automatically:
     • Hover tooltips showing data values
     • Zoom and pan capabilities
     • Ability to toggle legend items
     • Download options for saving the plot
3. Output Format Handling
   • We use knitr::is_html_output() to check if we’re rendering to HTML format.
   • For HTML output, we display the interactive plotly version.
   • For other formats (like PDF), we display the static ggplot2 version with a note about the interactive version.
   • This ensures the document works well in all output formats.

Results Interpretation

Interactive plots offer several advantages for ecological data exploration:

1. Data Exploration Benefits:
   • Detail on demand: Users can hover over points to see exact values without cluttering the visualization.
   • Selective viewing: Legend items can be clicked to show/hide specific groups (e.g., different species).
   • Focus on regions of interest: Zoom functionality allows detailed examination of specific data regions.
2. Ecological Applications:
   • Species trait analysis: Explore relationships between multiple morphological traits.
   • Environmental gradients: Investigate how species respond to environmental variables.
   • Outlier identification: Easily identify and examine unusual data points.
   • Data quality control: Interactive features help spot potential errors or anomalies.
3. Communication Advantages:
   • Engagement: Interactive plots engage readers more effectively than static images.
   • Self-guided exploration: Readers can investigate aspects of the data that interest them most.
   • Reduced complexity: Complex datasets can be presented more clearly when users can focus on specific elements.
4. Practical Considerations:
   • Interactive plots are only available in HTML output formats.
   • They may require more computational resources to render.
   • Always provide alternative static versions for PDF outputs or publications.
   • Consider accessibility for users who may rely on keyboard navigation.

PROFESSIONAL TIP: Creating Effective Interactive Visualizations

When using interactive plots in scientific communication:

1. Purpose-Driven Interactivity:
   • Add interactive elements with clear purpose, not just for novelty
   • Include only interactions that reveal meaningful patterns or details
   • Consider what specific questions users might want to answer through interaction
2. Performance and Compatibility:
   • Test interactive visualizations on different devices and browsers
   • Optimize for reasonable file sizes (especially for web deployment)
   • Provide static fallbacks for non-HTML formats and accessibility
3. Tooltip Design:
   • Include units of measurement in tooltips
   • Format numeric values appropriately (proper decimal places)
   • Include contextual information beyond just the raw values
   • Consider hierarchical information display (most important first)
4. Scientific Rigor:
   • Ensure interactive features don’t mislead or obscure statistical significance
   • Maintain appropriate axis scales during zooming
   • Include uncertainty or confidence intervals where applicable
   • Document interactive capabilities in figure captions or methods sections

7.3.2 Interactive Maps with leaflet

For spatial ecological data, interactive maps can be particularly useful:

library(leaflet)
library(ggplot2)
library(knitr)

# Create sample ecological site data
sites <- data.frame(
  name = c("Forest Reserve", "Wetland Study Area", "Grassland Transect",
           "Mountain Research Station", "Coastal Monitoring Site"),
  lat = c(37.7749, 37.8, 37.75, 37.85, 37.7),
  lng = c(-122.4194, -122.45, -122.5, -122.4, -122.3),
  habitat = c("Forest", "Wetland", "Grassland", "Alpine", "Coastal"),
  species_count = c(120, 85, 65, 95, 110)
)

# Create a color palette based on habitat type
habitat_colors <- c("darkgreen", "blue", "gold", "purple", "lightblue")
names(habitat_colors) <- c("Forest", "Wetland", "Grassland", "Alpine", "Coastal")

if (knitr::is_html_output()) {
  # For HTML output, create an interactive leaflet map
  habitat_pal <- colorFactor(
    palette = habitat_colors,
    domain = sites$habitat
  )

  # Create an interactive map
  leaflet(sites) %>%
    addProviderTiles("Esri.WorldTopoMap") %>%
    addCircleMarkers(
      ~lng, ~lat,
      color = ~habitat_pal(habitat),
      radius = ~sqrt(species_count) * 1.5,
      fillOpacity = 0.7,
      stroke = TRUE,
      weight = 1,
      popup = ~paste("<b>", name, "</b><br>",
                    "Habitat: ", habitat, "<br>",
                    "Species Count: ", species_count)
    ) %>%
    addLegend(
      position = "bottomright",
      pal = habitat_pal,
      values = ~habitat,
      title = "Habitat Type"
    )
} else {
  # For non-HTML output, create a static ggplot map
  ggplot(sites, aes(x = lng, y = lat, color = habitat, size = species_count)) +
    borders("world", regions = "USA", fill = "gray90") +
    geom_point(alpha = 0.7) +
    scale_color_manual(values = habitat_colors) +
    scale_size_continuous(range = c(3, 8)) +
    coord_fixed(1.3) +
    theme_minimal() +
    labs(
      title = "Ecological Study Sites",
      subtitle = "Note: Interactive version available in HTML output",
      x = "Longitude",
      y = "Latitude",
      color = "Habitat Type",
      size = "Species Count"
    )
}

Ecological study sites across different habitat types. In the HTML version, this map is interactive and allows zooming, panning, and clicking on markers for details.

Code Explanation

This code creates an interactive map of ecological study sites using the leaflet package. Let’s break down the key components:

1. Data Preparation
   • We create a sample dataset of ecological study sites with location coordinates (latitude and longitude), habitat types, and species counts.
   • We define a custom color palette that maps habitat types to ecologically intuitive colors (e.g., green for forest, blue for wetland).
2. Creating the Interactive Map
   • We use the leaflet() function to initialize a map with our site data.
   • We add a topographic base map with addProviderTiles("Esri.WorldTopoMap").
   • We represent each study site with a circle marker using addCircleMarkers():
     • The marker color corresponds to the habitat type.
     • The marker size is proportional to the square root of the species count (a common transformation for visual scaling).
     • We add popups that display site details when a marker is clicked.
   • We include a legend with addLegend() to explain the habitat type colors.
3. Output Format Handling
   • Similar to the previous example, we check if we’re rendering to HTML.
   • For non-HTML formats, we create a static map using ggplot2 as a fallback.
   • This ensures the document is useful in all output formats.

Results Interpretation

Interactive maps offer powerful capabilities for ecological spatial data visualization:

1. Spatial Data Exploration:
   • Context awareness: Base maps provide geographical context (topography, water bodies, urban areas).
   • Multi-scale examination: Zoom functionality allows viewing patterns at different spatial scales.
   • Detailed information: Popups provide detailed site-specific information without cluttering the map.
2. Ecological Applications:
   • Field site selection: Visualize potential study locations based on habitat types and existing data.
   • Biodiversity hotspots: Map species richness across different locations.
   • Sampling design: Plan spatially balanced sampling strategies.
   • Stakeholder engagement: Create accessible maps for communicating with non-specialists.
3. Analytical Advantages:
   • Spatial pattern recognition: Identify clustering or dispersion of ecological phenomena.
   • Environmental correlations: Overlay ecological data with environmental features.
   • Accessibility: Make complex spatial data accessible to broader audiences.
   • Data integration: Combine multiple spatial datasets in a single interactive interface.
4. Practical Considerations:
   • Interactive maps require HTML output format.
   • Consider data privacy when mapping sensitive locations (e.g., endangered species).
   • Ensure color choices are meaningful and accessible.
   • Balance information density with clarity.

PROFESSIONAL TIP: Best Practices for Ecological Mapping

When creating interactive maps for ecological data:

1. Base Map Selection:
   • Choose base maps appropriate to your ecological context (topographic for terrestrial studies, bathymetric for marine)
   • Consider offline capability for field use with packages like mapview
   • Be aware of licensing issues for base maps in publications
2. Data Representation:
   • Use ecologically intuitive color schemes (greens for forests, blues for aquatic)
   • Scale markers appropriately (√n transformation for count data often works well)
   • Consider using multiple layers for different data types (species, environmental variables)
   • Add scale bars and north arrows for static versions
3. Sensitive Location Handling:
   • Deliberately obscure precise locations of endangered species or sensitive habitats
   • Consider using grid cells, jittering, or reduced precision for protected species
   • Include appropriate disclaimers about location precision
   • Follow relevant regulations (e.g., IUCN guidelines for endangered species)
4. Technical Considerations:
   • Test on mobile devices when field use is anticipated
   • Optimize popup content for readability on small screens
   • Include data download capabilities when appropriate
   • Document the coordinate reference system (CRS) used

Interactive visualizations transform static ecological data into explorable resources, allowing readers to engage more deeply with your research findings and discover patterns that might otherwise remain hidden.

7.4 Network Visualizations

{{ … }}