07 – Plotting

Sebastian Raschka 6/17/2020

• Plotting
  • Stem-and-Leaf plots
  • Histograms
  • Barplot
  • Scatterplots
  • Boxplots

Source file: https://github.com/rasbt/R-notes/blob/master/07-plotting.Rmd

Plotting

• There are several different plotting libraries available in R. Here, we are going to focus on the built-in plotting capabilities (interested users may want to take a look at ggplot2, which is a popular plotting library some users prefer. You can find more information about it at https://ggplot2.tidyverse.org)

Stem-and-Leaf plots

• The example dataset used here contains information about the bark and concealment distances (in meters) of rural and urban prairie dogs:

• Stem-and-leaf plots ar probably the simplest type of plot, and we could construct it conveniently by hand for small dataset if we had to.

• using the stem function, we can construct the stem-and-leaf plot in R as followes.

urban_barkdist <- c(29, 10, 15, 41, 18, 18, 12, 45, 34, 30, 22, 26, 18)
urban_concealdist <- c(8, 8, 10, 8, 9, 10, 9, 10, 9, 6, 6, 7, 5)
rural_barkdist <- c(40, 47, 38, 59, 45, 52, 57, 50, 50, 49, 50, 43)
rural_concealdist <- c(14, 12, 24, 30, 37, 46, 39, 22, 38, 38, 20, 33)

stem(urban_barkdist)

## 
##   The decimal point is 1 digit(s) to the right of the |
## 
##   1 | 025888
##   2 | 269
##   3 | 04
##   4 | 15

• The left most digit (before the “|”) is the first digit of the values in the dataset. The values next to the stem represent the second digit of the values shown there. in the case of the urban_barkdist variable, we have 10, 12, 15, 18, 18, 18 (sorted in ascending order) for the first row, which become 1 | 025888
• This type of plot provides us with a quick view of the distribution of the data. In the example above, we have a right-skewed distribution (the tail is to the right.)

Histograms

hist(urban_barkdist)

hist(urban_barkdist, breaks=seq(0, 50, by = 5))

Barplot

barplot(c(mean(urban_barkdist), mean(rural_barkdist)),
        xlab="Bark distance (m)",
        names.arg=c("Urban", "Rural"),
        ylim=c(0, 65))

Scatterplots

plot(urban_barkdist, urban_concealdist,
     xlab="Bark distance (m)",
     ylab="Concealment discance (m)")

plot(urban_barkdist, urban_concealdist,
     xlim=range(1, 60), 
     ylim=range(1, 60),
     xlab='Bark distance (m)', 
     ylab='Concealment distance (m)');
points(rural_barkdist, rural_concealdist, pch=2)
legend(1, 50, pch=1:2, legend=c("Urban", "Rural"))

Boxplots