Data Visualization

Topics: ggplot2, visualization

Visualizations help reveal patterns in clinical data — such as disease prevalence, lab result distributions, and relationships between vital signs — and effectively communicate findings to support medical decisions or publications.

Basic Plotting with ggplot2

Load package and set default theme

library(ggplot2)
theme_pub <- theme_classic(base_size = 35) +
  theme(
    plot.title = element_text(
      size = 20,
      face = "bold",
      hjust = 0.5
    ),
    axis.title = element_text(
      size = 20,
      face = "bold"
    ),
    axis.text = element_text(
      size = 20,
      color = "black"
    ),
    legend.title = element_text(
      size = 20,
      face = "bold"
    ),
    legend.text = element_text(
      size = 20
    ),
    plot.margin = margin(10, 10, 10, 10)
  )

Histogram — Distribution of Glucose Levels

💡 Use case: Quickly visualize how many patients fall into normal, prediabetic, or diabetic glucose ranges.

ggplot(df, aes(x = Glucose)) +
  geom_histogram(
    bins = 20,
    fill = "#2C7FB8",
    color = "white",
    linewidth = 0.5
  ) +
  labs(
    title = "Distribution of Fasting Glucose Levels",
    x = "Glucose (mg/dL)",
    y = "Number of Patients"
  ) +
  theme_pub

Box plot — Glucose by Diagnosis

💡 Use case: Compare glucose variability across different conditions (e.g., Diabetes, Hypertension, Normal).

ggplot(df, aes(x = Diagnosis, y = Glucose, fill = Diagnosis)) +
  geom_boxplot(
    width = 0.5,
    alpha = 0.8,
    outlier.shape = NA
  ) +
  geom_jitter(
    width = 0.15,
    alpha = 0.5,
    size = 1.8
  ) +
  scale_fill_brewer(
    palette = "Set2",
    guide = "none"
  ) +
  labs(
    title = "Glucose Distribution by Diagnosis",
    x = "Diagnosis",
    y = "Glucose (mg/dL)"
  ) +
  theme_pub +
  theme(
    axis.text.x = element_text(
      angle = 0,
      hjust = 0.5
    )
  )

Advanced Plotting with ggplot2

Here, we’ll produce the following plots:

Correlation heatmap
Scatter plot
Density plot
Violin plot
Count/bar plot

Correlation heatmap

💡 Use case: Identify correlations (e.g., between Age and Blood Pressure).

library(reshape2)   # or use tidyr::pivot_longer

corr_matrix <- cor(df |> select(Age, Glucose, Blood_Pressure), use = "complete.obs")
corr_melt   <- melt(corr_matrix)

ggplot(corr_melt,
       aes(
         x = Var1,
         y = Var2,
         fill = value
       )) +
  geom_tile(
    color = "white",
    linewidth = 0.5
  ) +
  geom_text(
    aes(label = sprintf("%.2f", value)),
    size = 4.5
  ) +
  scale_fill_gradient2(
    low = "#2166AC",
    mid = "white",
    high = "#B2182B",
    midpoint = 0,
    limits = c(-1, 1),
    name = "Correlation"
  ) +
  coord_fixed() +
  labs(
    title = "Correlation Matrix of Clinical Variables",
    x = NULL,
    y = NULL
  ) +
  theme_minimal(base_size = 14) +
  theme(
    plot.title = element_text(
      face = "bold",
      size = 16,
      hjust = 0.5
    ),
    axis.text = element_text(
      color = "black"
    ),
    panel.grid = element_blank()
  )

Scatter plot with regression line

💡 Use case: Explore whether glucose tends to rise with age and how it differs by diagnosis.

ggplot(
  df,
  aes(
    x = Age,
    y = Glucose,
    color = Diagnosis
  )
) +
  geom_point(
    size = 2.5,
    alpha = 0.7
  ) +
  geom_smooth(
    method = "lm",
    se = TRUE,
    linewidth = 1
  ) +
  labs(
    title = "Association Between Age and Glucose",
    x = "Age (Years)",
    y = "Glucose (mg/dL)",
    color = "Diagnosis"
  ) +
  theme_pub

Density plot

💡 Use case: Compare glucose distributions across diagnoses using kernel density estimation.

ggplot(
  df,
  aes(
    x = Glucose,
    fill = Diagnosis,
    color = Diagnosis
  )
) +
  geom_density(
    alpha = 0.25,
    linewidth = 1
  ) +
  labs(
    title = "Distribution of Glucose Levels by Diagnosis",
    x = "Glucose (mg/dL)",
    y = "Density"
  ) +
  theme_pub

density plot of glucose levels by diagnosis

Violin plot

💡 Use case: Examine BMI variation among diagnostic categories.

ggplot(
  df,
  aes(
    x = Diagnosis,
    y = Blood_Pressure,
    fill = Diagnosis
  )
) +
  geom_violin(
    trim = FALSE,
    alpha = 0.7
  ) +
  geom_boxplot(
    width = 0.12,
    fill = "white",
    linewidth = 0.5
  ) +
  labs(
    title = "Blood Pressure Distribution by Diagnosis",
    x = "Diagnosis",
    y = "Blood Pressure (mmHg)"
  ) +
  scale_fill_brewer(
    palette = "Set2",
    guide = "none"
  ) +
  theme_pub +
  theme(
    axis.text.x = element_text(
      angle = 45,
      hjust = 1
    )
  )

violiin plot of blood pressure distribution by diagnosis

Count/bar plot

💡 Use case: Display patient counts per glucose category within each diagnosis.

ggplot(
  df,
  aes(
    x = Glucose_Category,
    fill = Diagnosis
  )
) +
  geom_bar(
    position = position_dodge(width = 0.8),
    width = 0.7
  ) +
  labs(
    title = "Glucose Categories by Diagnosis",
    x = "Glucose Category",
    y = "Count",
    fill = "Diagnosis"
  ) +
  theme_pub +
  theme(
    axis.text.x = element_text(
      angle = 0,
      hjust = 0.5
    )
  )

count bar chart of glucose categories by diagnosis

Combining Multiple Plots

💡 Use case: Summarize proportions of diagnoses and visualize clinical indicators together.

library(patchwork)

p1 <- ggplot(df, aes(x = Age)) +
  geom_histogram(
    bins = 15,
    fill = "#2C7FB8",
    color = "white"
  ) +
  labs(
    title = "A. Age Distribution",
    x = "Age (Years)",
    y = "Count"
  ) +
  theme_pub

avg_glucose <- df |>
  group_by(Diagnosis) |>
  summarise(
    Mean_Glucose = mean(
      Glucose,
      na.rm = TRUE
    ),
    .groups = "drop"
  ) |>
  arrange(desc(Mean_Glucose))

p2 <- ggplot(
  avg_glucose,
  aes(
    x = reorder(Diagnosis, Mean_Glucose),
    y = Mean_Glucose
  )
) +
  geom_col(
    fill = "#2C7FB8",
    width = 0.7
  ) +
  coord_flip() +
  labs(
    title = "B. Mean Glucose by Diagnosis",
    x = NULL,
    y = "Mean Glucose (mg/dL)"
  ) +
  theme_pub

p3 <- ggplot(
  df,
  aes(
    x = Age,
    y = Blood_Pressure
  )
) +
  geom_point(
    alpha = 0.7,
    size = 2
  ) +
  geom_smooth(
    method = "lm",
    color = "black",
    se = TRUE
  ) +
  labs(
    title = "C. Age vs Blood Pressure",
    x = "Age (Years)",
    y = "Blood Pressure (mmHg)"
  ) +
  theme_pub

p4 <- ggplot(
  df,
  aes(
    x = "",
    fill = Diagnosis
  )
) +
  geom_bar(
    width = 1
  ) +
  coord_polar("y") +
  labs(
    title = "D. Diagnosis Composition"
  ) +
  theme_void(base_size = 14) +
  theme(
    plot.title = element_text(
      face = "bold",
      size = 16,
      hjust = 0.5
    )
  )

(p1 + p2) /
(p3 + p4)

Key Takeaways

ggplot2 uses a layered grammar of graphics — start with data, add aesthetics, then layers.
Plots like density, violin, and scatter with regression add statistical depth.
Visualizations are crucial for pattern recognition, data quality assessment, and research presentation.
Use them to compare distributions, examine relationships, and summarize populations in any data.

Data Filtering and Transformation Statistical Analysis