Chapter 4 Plotting using one variable at a time

4.1 Histograms- plots the distribution of a numerical variable

Note the customizations of the histogram: 1. we colored the bars pretty 2. ‘alpha’ adds transparency to the object, useful when you have overlapping objects, goes from 0 (transparent) to 1 (opaque)

library(ggplot2)

ggplot(ChickWeight, aes(weight))+
  geom_histogram(fill='cyan4',color='black',alpha=0.5)+
  theme_minimal(base_size = 20)+
  ylab("Frequency")+ xlab("Weight (g)")

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Figure 4.1: Histogram

Improving ‘binwidth’ value, it is the number of categories data is divided into and usually defaults to 30, but we can go up and down, let’s go to 50 here

library(ggplot2)

ggplot(ChickWeight, aes(weight))+
  geom_histogram(fill='cyan4',color='black',alpha=0.5, binwidth = 50)+
  theme_minimal(base_size = 20)+
  ylab("Frequency")+ xlab("Weight (g)")

Figure 4.2: Histogram

4.2 Bar plots

Plot the distribution of a categorical variable

Here we will plot number of chicks following a particular diet.

library(ggplot2)

ggplot(ChickWeight, aes(x= Diet))+
  geom_bar(color='orange', fill='lavender')+
  theme_minimal(base_size = 20)+
  ylab("Count")+ xlab("Diet")

Figure 4.3: Bar plot

4.3 Ordering based on number of counts from lowest to highest

Let’s bring the magic of dplyer to do so

So we reorganize the data to calcualte number of chicks on each diet type

library(tidyverse)
reord.chick<-ChickWeight %>% 
 count(Diet) %>% arrange(n)
reord.chick

##   Diet   n
## 1    4 118
## 2    2 120
## 3    3 120
## 4    1 220

Here we plot using geom_col function where heights represents the value of the data and requires y aesthetics. We can use geom_text function to add the label to each column and use vjust to move the labels up and down

library(ggplot2)

ggplot(reord.chick, aes(x= reorder(Diet,n), y=n))+
  geom_col(color='orange', fill='lavender')+
  geom_text(aes(label=n), vjust=-0.7)+
  theme_minimal(base_size = 20)+
  ylab("Count")+ xlab("Diet")

Figure 4.4: Bar plot with labels

4.4 Adding percentages

Here we also get more adventerous with data wrangling and create new columns and plot them within the same pipe

ChickWeight %>% 
  count(Diet) %>% arrange(n) %>% 
mutate(percent = n / sum(n),
         percentlabel = paste0(round(percent*100), "%")) %>% 

ggplot( aes(x= reorder(Diet,percent), y=percent))+
  geom_col(color='orange', fill='lavender')+
  geom_text(aes(label=percentlabel), vjust=-0.7)+
  theme_minimal(base_size = 20)+
  ylab("Count")+ xlab("Diet")

Figure 4.5: Bar plot with percent labels

4.5 pie chart

library(ggplot2)

ggplot(reord.chick, aes(x= reorder(Diet,n), y=n,fill=Diet))+
  geom_col(width = 1, 
           stat = "identity", 
           color = "black") +
  geom_text(aes(label=n), vjust=-0.7)+
  theme_minimal(base_size = 20)+
  coord_polar("y", 
              start = 0,direction = -1 
              ) +
  theme_void() +
  ylab("Count")+ xlab("Diet")

## Warning: Ignoring unknown parameters: stat

Figure 4.6: Pie chart with labels