COSMOS logo

Visualizing Data

Zhaoxia Yu

Reminder

  • Close all apps on your computer.
  • Open slides for this session from the COSMOS website.

Preparation

  1. Load package named ‘ggplot2’ that used for visualization

  2. Load tidyverse in order to refresh data wrangling

  3. Load alzheimer_data

install/load ggplot2 and tidyverse

Code 
#install.packages('ggplot2')
library(ggplot2)

#install.packages('tidyverse')
library(tidyverse)

load alzheimer_data

Code 
# change the path to your personal folder path
alzheimer_data <- read.csv("../data/alzheimer_data.csv")

Data

Code 
glimpse(alzheimer_data)
Rows: 2,700
Columns: 57
$ id        <chr> "S060833", "S932623", "S755478", "S852291", "S011143", "S069…
$ diagnosis <int> 0, 0, 0, 0, 1, 0, 0, 2, 0, 2, 0, 0, 0, 1, 0, 1, 2, 2, 2, 1, …
$ age       <int> 74, 56, 77, 74, 75, 72, 64, 78, 73, 81, 66, 65, 66, 73, 78, …
$ educ      <int> 12, 16, 18, 20, 14, 16, 16, 17, 18, 13, 16, 16, 17, 20, 13, …
$ female    <int> 0, 1, 1, 1, 0, 1, 1, 0, 1, 0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, …
$ height    <dbl> 65.0, 62.0, 65.0, 62.0, 62.0, 61.8, 60.0, 69.0, 65.0, 71.0, …
$ weight    <int> 233, 110, 137, 112, 127, 141, 124, 152, 131, 197, 134, 144, …
$ bpsys     <int> 148, 110, 144, 120, 145, 107, 112, 134, 122, 120, 150, 126, …
$ bpdias    <int> 100, 75, 60, 60, 61, 65, 70, 74, 60, 70, 85, 78, 60, 72, 80,…
$ hrate     <int> 72, 60, 64, 72, 58, 83, 76, 70, 60, 76, 60, 60, 76, 60, 68, …
$ cdrglob   <dbl> 0.5, 0.0, 0.0, 0.0, 0.5, 0.0, 0.0, 0.5, 0.0, 1.0, 0.0, 0.0, …
$ naccgds   <int> 5, 1, 0, 0, 4, 1, 2, 0, 0, 5, 0, 1, 0, 0, 0, 6, 3, 1, 3, 4, …
$ delsev    <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
$ hallsev   <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
$ agitsev   <int> 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 0, 0, 0, 1, 0, 0, 0, 0, …
$ depdsev   <int> 2, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 2, 0, 0, 0, 3, 0, 0, 0, 0, …
$ anxsev    <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 2, 0, …
$ elatsev   <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
$ apasev    <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 3, 0, 1, 0, 0, 0, 2, 1, 0, 0, 0, …
$ disnsev   <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, …
$ irrsev    <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, 0, …
$ motsev    <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
$ nitesev   <int> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, …
$ appsev    <int> 0, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
$ bills     <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 8, 1, 3, 2, …
$ taxes     <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 1, 8, 2, 3, 3, …
$ shopping  <int> 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 2, 1, …
$ games     <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 8, 0, 0, 8, 0, 0, 8, 3, 0, 1, 0, …
$ stove     <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, …
$ mealprep  <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 8, 0, 0, 1, 3, 8, 0, 0, …
$ events    <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 3, 0, 0, 2, …
$ payattn   <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 2, 1, 1, 1, …
$ remdates  <int> 1, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 3, 0, 1, 2, …
$ travel    <int> 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 3, 0, 3, 1, …
$ naccmmse  <int> 30, 29, 30, 30, 27, 30, 30, 26, 29, 28, 30, 30, 29, 30, 29, …
$ memunits  <int> 8, 17, 19, 11, 6, 16, 11, 6, 4, 3, 14, 9, 13, 17, 14, 12, 9,…
$ digif     <int> 7, 11, 7, 6, 8, 7, 8, 9, 4, 7, 12, 9, 6, 10, 7, 6, 1, 8, 5, …
$ animals   <int> 17, 25, 19, 23, 19, 14, 28, 14, 14, 16, 16, 21, 11, 14, 21, …
$ traila    <int> 49, 16, 38, 54, 40, 44, 31, 69, 53, 34, 46, 27, 22, 38, 38, …
$ trailb    <int> 130, 47, 83, 100, 67, 100, 55, 168, 123, 90, 118, 72, 47, 85…
$ naccicv   <dbl> 1389.520, 1366.945, 1367.420, 1359.850, 1367.420, 1240.390, …
$ csfvol    <dbl> 381.840, 366.622, 343.176, 332.880, 390.415, 345.600, 310.43…
$ lhippo    <dbl> 2.2900, 3.2606, 2.6990, 3.0600, 2.9342, 3.2100, 3.6800, 1.73…
$ rhippo    <dbl> 2.9200, 3.3321, 2.5028, 3.0000, 3.2890, 3.0600, 3.8200, 2.30…
$ frcort    <dbl> 160.570, 187.874, 163.214, 165.120, 149.138, 140.220, 195.24…
$ lparcort  <dbl> 46.7100, 43.1023, 40.1172, 46.0700, 40.2664, 42.9700, 48.740…
$ rparcort  <dbl> 47.7300, 43.8414, 38.2377, 46.3700, 42.1025, 43.1300, 48.240…
$ ltempcor  <dbl> 57.9700, 60.3437, 58.8357, 54.2100, 57.3215, 55.2600, 66.770…
$ rtempcor  <dbl> 58.5800, 58.7091, 51.5753, 56.1600, 54.4138, 51.6100, 60.530…
$ lcac      <dbl> 3.3200, 3.7060, 3.2748, 2.6300, 3.8628, 2.2600, 2.7800, 2.98…
$ rcac      <dbl> 1.9800, 2.1906, 1.7054, 1.4400, 1.5277, 1.9300, 1.8600, 1.36…
$ lent      <dbl> 3.2000, 3.6755, 3.6207, 4.3300, 4.2328, 3.8200, 4.5000, 2.73…
$ rent      <dbl> 3.7300, 4.6463, 2.5787, 4.1000, 4.4572, 3.4900, 4.3700, 2.58…
$ lparhip   <dbl> 3.5800, 3.5534, 3.7515, 3.6000, 3.7079, 4.0700, 5.1100, 3.63…
$ rparhip   <dbl> 3.6800, 4.1952, 3.6703, 3.9200, 3.4988, 4.0100, 5.1300, 3.12…
$ lposcin   <dbl> 3.7500, 3.9091, 3.8686, 3.4500, 3.1321, 3.4500, 4.3700, 4.38…
$ rposcin   <dbl> 3.4400, 4.2362, 3.7062, 3.5300, 2.9051, 2.9200, 4.1800, 3.85…

ggplot

  • ggplot is based on grammar of graphics.
  • It is a powerful and flexible system for creating a wide range of visualizations.

3 steps of making a basic ggplot

1.Pick data

2.Map data onto aesthetics

3.Add the geometric layer

Visualizing a Categorical Variable

Let’s check the number of patients in each diagnosis group.

Step 1: Data

Code 
ggplot(data = alzheimer_data)

Step 2: Map to Aesthetics

Code 
ggplot(data = alzheimer_data,
       aes(x = diagnosis))

Step 3: Add Geometric Layer with +

Code 
ggplot(data = alzheimer_data,
       aes(x = diagnosis)) +
  geom_bar()

Additional Features

  1. Use pipe operator to feed the data to ggplot.
Code 
alzheimer_data %>% 
  ggplot(aes(x = diagnosis)) +
  geom_bar()

Additional Features

  1. Use factor() to add informative label to categorical variable (diagnosis)

Note: We will learn how to add labels to the axis later. Ignore labs(), theme_minmial() for now.

Code 
# label diagnosis 0, 1, 2 to healthy, impaired, and Alzheimer's
alzheimer_data <- alzheimer_data %>%
  mutate(diagnosis_name = factor(diagnosis, 
                            levels = c(0, 1, 2), 
                            labels = c("Healthy", "Impaired", "Alzheimer's"))) 

alzheimer_data %>% 
  ggplot(aes(x = diagnosis_name)) +
  geom_bar() +
  labs(
    title = "Count of Diagnosis Groups",
    x = "Diagnosis Type",
    y = "Count"
  )  +  
  theme_minimal()

Categorical by Categorical Variable

We can also use bar plot to visualize how the count of one categorical variable (e.g. gender) are distributed across the levels of another (e.g. diagnosis).

Categorical by Categorical Variable

  1. Stacked Bar Plot (Counts)
Code 
# label female 0, 1 to male, female
alzheimer_data = alzheimer_data %>% 
         mutate(female = factor(
           female, 
           levels = c(0,1), 
           labels = c("Male", "Female")))
                
alzheimer_data %>%  
  ggplot(aes(x = female, fill = diagnosis_name)) +
  geom_bar(position = "stack") +
  labs(
    title = "Count of Diagnosis by Gender",
    x = "Gender",
    y = "Count",
    fill = "Diagnosis"
  ) +
  theme_minimal()

Categorical by Categorical Variable

  1. Grouped (Dodged) Bar Plot (Counts)
Code 
ggplot(alzheimer_data, aes(x = female, 
                          fill = diagnosis_name)) +
  geom_bar(position = "dodge") +
  labs(
    title = "Count of Diagnosis by Gender (Grouped)",
    x = "Gender",
    y = "Count",
    fill = "Diagnosis"
  ) +
  theme_minimal()

Categorical by Categorical Variable

  1. Proportions (Within Groups)
Code 
alzheimer_data %>% 
  ggplot(aes(x = female,fill = diagnosis_name)) +
  geom_bar(position = "fill") +  # Normalizes to 100% per group
  labs(
    title = "Proportion of Diagnosis by Gender",
    x = "Gender",
    y = "Proportion",
    fill = "Diagnosis"
  ) + 
  theme_minimal()

Visualize a Numerical Variable

  • What data type is the volume of left hippocampus (lhippo) ?
  • What is the distribution of the observed volume of left hippocampus?

Visualize a Numerical Variable: boxplot

Code 
ggplot(data= alzheimer_data,
       mapping= aes(x= lhippo)) +
  geom_boxplot() + 
  labs(title="Boxplot of lhippo",
      x="Left Hippocampal Volume (cm³)") +  
  theme_minimal()

Visualize a Numerical Variable: histogram

Code 
ggplot(data= alzheimer_data,
       mapping= aes(x= lhippo)) +
  geom_histogram(bins = 30) +
  labs(
    title="Violin of lhippo",
    x="Left Hippocampal Volume (cm³)",
    y="Count") +  
  theme_minimal()

Visualize a Numerical Variable by a Categorical Variable

Code 
ggplot(data = alzheimer_data,
       mapping = aes(
         x = female,  
         y = lhippo,
         fill = female)  # Optional: color by gender
       ) +
  geom_boxplot() +
  labs(
    title = "Boxplot of Left Hippocampal Volume by Gender",
    x = "Gender",
    y = "Left Hippocampal Volume (cm³)",  
    fill = "Gender"  # Legend title
  ) +
  theme_minimal()

Visualize a Numerical Variable by a Categorical Variable

Code 
ggplot(data = alzheimer_data,
       mapping = aes(
         x = lhippo,
         fill =female)  # Ensure 'female' is a labeled factor
       ) +
  geom_histogram(bins = 35, alpha = 0.7) +  # Adjust transparency  
  labs(
    title = "Histogram of Left Hippocampal Volume by Gender",
    x = "Left Hippocampal Volume (cm³)",  
    y = "Count",
    fill = "Gender"  # Legend title
  ) +
  theme_minimal() +
  scale_fill_manual(values = c("Male" = "skyblue", "Female" = "salmon"))  # Custom colors (optional)

Visualize Numerical Variable by Numerical Variable

Will be covered in next session.

Summary

Practice

  • Visualize lhippo by diagnosis
  • Visualize lhippo by gender in healthy subjects