Examples on using functions from ggplot2 and plotly

Overview

This R Markdown file contains sample codes for different types of plots that can be generate with ggplot2 package.

Set up

Import packages

Histogram

1. Regular histogram

• binwidth: set the width of the bins
• bins: set the number of bins in the histogram
• alpha: controls transparency. 0 means no fill color, 1 means solid fill color
• color: color for the edges of the histogram
• fill: filling color in the histogram

# create a histogram with binwidth 0.1, color the edge of the graph with red, color the inside of the histogram with pink, alpha = 0.5 means there's some transparency 
movies %>% 
ggplot(aes(x = rating)) +
  geom_histogram(binwidth = 0.1, color = 'red', fill = 'pink', alpha = 0.5 ) 

2. Better histogram

use count to color the histogram

library(ggplot2movies)
library(dplyr)

movies %>% 
ggplot(aes(x = rating)) +
  geom_histogram(binwidth = 0.1, aes(fill = ..count..)) 

Scatter plot

1. Basic Scatter plot

• color can also be defined using hex color like ‘#33CCCC’, check https://www.color-hex.com/

mtcars %>% 
  ggplot(aes(x = wt, y = mpg))+
  geom_point(size = 5, color = '#33CCCC')

2. Change features of the points with variables in Scatter plot

Example 1.1: change the size of the points with variables in Scatter plot

mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point(aes(size = hp))

Example 1.2: Note:use factor() because cyl is not continues. cyl contains only even numbers (4,6,8)

mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point(aes(size = factor(cyl)))

Example 2.1: change the shape and color of the points with variables

mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point(aes(shape = factor(cyl), color = factor(cyl)), size = 4)

Example 2.2: scale_color_gradient sets the colors for the points based on the value of hp

mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point(aes(color = hp),size = 5) + 
  scale_color_gradient(low = 'blue', high = 'red')

Example 3: geom_smooth adds a smooth line that fits the points

txhousing %>% 
  ggplot(aes(x = sales, y = volume)) +
  geom_point(aes(color = sales), alpha = 0.6) +
  scale_color_gradient(low = 'blue', high = 'red')+
  geom_smooth(color = 'green')

Bar plot

Example 1:

mpg %>% 
  ggplot(aes(x = class)) +
  geom_bar( color = 'blue', fill = 'blue')

Example 2:

mpg %>% 
  ggplot(aes(x = hwy)) +
  geom_bar(fill = '#F59D92')

Example 3: use fill = drv to add more information to the bar plot;

mpg %>% 
  ggplot(aes(x = class)) +
  geom_bar(aes(fill = drv))

Example 4: use coord_flip to switch x and y axis.

mpg %>% 
  ggplot(aes(x = manufacturer)) +
  geom_bar(aes(fill = factor(cyl))) +
  coord_flip()

Example 5: position = 'dodge' gives multiple bar plot side by side instead of stacking

mpg %>% 
  ggplot(aes(x = class)) +
  geom_bar(aes(fill = drv), position = 'dodge') 

Example 6: position = 'fill' shows the percentage of each class

mpg %>% 
  ggplot(aes(x = class)) +
  geom_bar(aes(fill = drv), position = 'fill') 

Boxplot

Example 1: geom_bocplot makes box plot

mtcars %>% 
  ggplot(aes(x = factor(cyl), y = mpg)) +
  geom_boxplot()

Example 2:

mtcars %>% 
  ggplot(aes(x = factor(cyl), y = mpg)) +
  geom_boxplot() +
  coord_flip()

Example 3: use theme_dark() to change the background

mtcars %>% 
  ggplot(aes(x = factor(cyl), y = mpg)) +
  geom_boxplot(aes(fill = factor(cyl))) +
  theme_dark()

Heat map

Example 1: geom_bin2d() makes regular heat map

movies %>% 
  ggplot(aes(x = year, y = rating)) +
  geom_bin2d()

Example 2:

movies %>% 
  ggplot(aes(x = year, y = rating)) +
  geom_bin2d() + scale_fill_gradient(high = 'red', low = 'green')

Example 3:

movies %>% 
  ggplot(aes(x = year, y = rating)) +
  geom_bin2d(binwidth = c(3,1)) + scale_fill_gradient(high = 'red', low = 'blue')

Example 4: geom_hex() makes heat map with hexagons

movies %>% 
  ggplot(aes(x = year, y = rating)) +
  geom_hex() + scale_fill_gradient(high = 'red', low = 'blue')

density plot

movies %>% 
  ggplot(aes(x = year, y = rating)) +
  geom_density2d() 

movies %>% 
  ggplot(aes(x = year, y = rating)) +
  geom_point()

Coordanates Faceting

Example 1: use coord_cartesian(xlim = c(1, 4), ylim = c(15,30)) to set the limits for x and y axis

mpg %>% 
  ggplot(aes(x = displ, y = hwy)) +
  geom_point() +
  coord_cartesian(xlim = c(1, 4), ylim = c(15,30))

Example 2: change the length of axis with coord_fixed

mpg %>% 
  ggplot(aes(x = displ, y = hwy)) +
  geom_point() +
  coord_fixed(ratio = 1/3)# ratio = x/y

Example 3: facet_grid(.~cyl) create multiple plots for different cyl values. .~ cyl split on x-axis

mpg %>% 
  ggplot(aes(x = displ, y = hwy)) +
  geom_point() +
  facet_grid(.~cyl)

Example 4: facet_grid(drv~.) create multiple plots for different cyl values. drv~. split on y-axis Note: .~var_name split on x-axis and var_name~. split on y-axis

mpg %>% 
  ggplot(aes(x = displ, y = hwy)) +
  geom_point() +
  facet_grid(drv~.)

Example 5: facet_grid(y~x) split on both y-axis and x-axis

mpg %>% 
  ggplot(aes(x = displ, y = hwy)) +
  geom_point() +
  facet_grid(drv~cyl)

Themes

Example 1:

• theme_set(theme_minimal()) set the themes for all the plots

theme_set(theme_minimal())
mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point()

Example 2: theme_dark()

mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point() +
  theme_dark()

Example 3: load the library ggthemes and use ?ggthemes to find out more themes

library(ggthemes) 
mtcars %>% 
  ggplot(aes(x = wt, y = mpg)) +
  geom_point() + 
  theme_economist()

Exercise

df <- read.csv('Economist_Assignment_Data.csv')
head(df)

df %>% 
  ggplot(aes(x = CPI, y = HDI)) +
  geom_point(aes(color = Region))

df %>% 
  ggplot(aes(x = CPI, y = HDI)) +
  geom_point(aes(color = Region), shape = 'o', size = 3)

df %>% 
  ggplot(aes(x = CPI, y = HDI)) +
  geom_point(aes(color = Region), shape = 'o', size = 3) +
  geom_smooth(aes(group=1))

df %>% 
  ggplot(aes(x = CPI, y = HDI)) +
  geom_point(aes(color = Region), shape = 'o', size = 3) +
  geom_smooth(aes(group = 1),method = 'lm', formula = y ~ log(x), se = FALSE, color = 'red')

df %>% 
  ggplot(aes(x = CPI, y = HDI)) +
  geom_point(aes(color = Region), shape = 'o', size = 3) +
  geom_smooth(aes(group=1),method = 'lm', formula = y ~ log(x),
              se = FALSE, color = 'red') +
  geom_text(aes(label = Country, color = Region)) 

pointsToLabel <- c("Russia", "Venezuela", "Iraq", "Myanmar", "Sudan", "Afghanistan", "Congo", "Greece", "Argentina", "Brazil", "India", "Italy", "China", "South Africa", "Spane", "Botswana", "Cape Verde", "Bhutan", "Rwanda", "France", "United States", "Germany", "Britain", "Barbados", "Norway", "Japan", "New Zealand", "Singapore")

df %>% 
  ggplot(aes(x = CPI, y = HDI)) +
  geom_point(aes(color = Region), shape = 'o', size = 3) +
  geom_smooth(method = 'lm', formula = y ~ log(x),
              se = FALSE, color = 'red') +
   geom_text(aes(label = Country), color = "gray20", 
                data = subset(df, Country %in% pointsToLabel),check_overlap = TRUE) +
  theme_bw() +
  scale_x_continuous(name = "Corruption Perceptions Index, 2011 (10=least corrupt)",limits = c(.9, 10.5),breaks=1:10) +
  scale_y_continuous(name = "Human Development Index, 2011 (1=Best)",limits = c(0.2, 1.0)) +
  ggtitle("Corruption and Human development") +
  theme_economist_white()

Label only a part of data

1. Create a vector that contains all the labels to be showed on the plot, like pointsToLabel.
2. In geom_text(), use data = subset(data_name, variable_name %in% vector_name, check_overlap = TRUE)

pointsToLabel <- c("Russia", "Venezuela", "Iraq", "Myanmar", "Sudan", "Afghanistan", "Congo", "Greece", "Argentina", "Brazil", "India", "Italy", "China", "South Africa", "Spane", "Botswana", "Cape Verde", "Bhutan", "Rwanda", "France", "United States", "Germany", "Britain", "Barbados", "Norway", "Japan", "New Zealand", "Singapore")

df %>% 
  ggplot(aes(x = CPI, y = HDI)) +
  geom_point(aes(color = Region), shape = 'o', size = 3) +
  geom_smooth(method = 'lm', formula = y ~ log(x),
              se = FALSE, color = 'red') +
   geom_text(aes(label = Country), color = "gray20", 
                data = subset(df, Country %in% pointsToLabel),check_overlap = TRUE) +
  theme_bw() +
  scale_x_continuous(name = "Corruption Perceptions Index, 2011 (10=least corrupt)",limits = c(.9, 10.5),breaks=1:10) +
  scale_y_continuous(name = "Human Development Index, 2011 (1=Best)",limits = c(0.2, 1.0)) +
  ggtitle("Corruption and Human development") +
  theme_economist_white()

Plotly

To use plotly to create interactive plots, simply add ggplotly() around the code.

https://plotly.com/r/

ggplotly(mtcars %>%
  ggplot(aes(mpg ,wt)) +
  geom_point())