In jr-packages/jrIntroduction: Jumping Rivers: Introduction to R

Base Graphics

data(movies, package = "ggplot2movies")
sub_movies = movies[1:500, ]

op = par(mfrow = c(2, 1), mar = c(3, 3, 3, 1))
plot(sub_movies$length, sub_movies$rating)
text(0, 8.5, "(a)")

plot(sub_movies$length, sub_movies$rating,
     ylab = "Rating", xlab = "Movie Length",
     ylim = c(0, 10), xlim = c(0, 250),
     pch = 21, bg = sub_movies$Action + 1,
     panel.first = grid(), cex = 0.7)
text(0, 9.45, "(b)")
par(op)

The standard plot

The standard plotting command in R is plot(). It takes two coordinate arguments: plot(x, y).^[Actually, you can just type plot(y) and it will generate a default for the x-axis] For example,

data(movies, package = "ggplot2movies")
sub_movies = movies[1:500, ]
## Plot first 500 movies
plot(sub_movies$length, sub_movies$rating)

gives figure \@ref(fig:moviesscatter) a. The basic plot command has multiple arguments:

• main, xlab and ylab control the axis labels.
• xlim and ylim control the axis ranges.
• lwd controls the thickness of the lines.
• lty specifies the line type, e.g. dotted lines.
• type = 'p' - draws points.
• type = 'l' - draws lines.
• type = 'b' - draws lines and points.
• type = 'o' - draws lines and points over-plotted.
• log = 'y' - use the log scale on the y (or x) axis.

Using the default settings is fine for investigating data, but we often have to set limits and colours explicitly:

plot(sub_movies$length, sub_movies$rating,
     ylab = "Rating", xlab = "Movie Length",
     ylim = c(0, 10), xlim = c(0, 250),
     pch = 21, bg = sub_movies$Action + 1,
     panel.first = grid(), cex = 0.7)

to get figure \@ref(fig:moviesscatter) b. To add lines, we use the lines() function. Horizontal and vertical lines can be added using the convenience function abline(), viz.

## Add a horizontal line at the origin
abline(h = 0)
## Vertical line
abline(v = 1)

and plain text is added using the text() function:

## Adds "(b)" to (0, 9.3) on the plot
text(0, 9.3, "(b)")

Bar charts

The most useful way to display qualitative data is usually with a bar chart. The length of each bar is equal to the frequency of the corresponding value of the variable in the sample of data. Note that the widths of the bars should be equal to avoid giving a false impression.

Figure \@ref(fig:moviesbarchart) shows the breakdown in MPAA ratings for the movie data set. To create a bar chart in R we use the following commands:

op = par(mgp = c(2.5, 0.5, 0), mar = c(3.5, 3.5, 2, 1))
barplot(table(movies$mpaa), xlab = "MPAA Rating",
        ylab = "Frequency", border = "black")
par(op)

Histograms

op = par(mfrow = c(2, 1), mar = c(3, 3, 3, 1))
hist(movies$rating, col = "grey", main = "User rating",
     freq = TRUE, xlim = c(1, 10), xlab = "Rating")
hist(movies$budget, col = "grey", main = "Film budget",
     freq = FALSE, xlab = "Budget")
par(op)

To represent the distribution of a sample of values of a continuous variable we can use a histogram. The range of values of the variable is divided into intervals, known as classes, and the frequencies in classes are represented by columns. As the variable is continuous, there are no gaps between neighbouring columns, unlike a bar chart. Note also that, strictly speaking, it is the area of the column which is equal to the frequency, not the height. The reason for this is that columns need not be of the same width. Computer software tends to use columns of the same width. However this default can be overridden in R if you really want to. Figure \@ref(fig:movieshist) shows histograms of film ratings and budgets.

\indent To generate Figure \@ref(fig:movieshist) in R we use the following commands:

hist(movies$rating, col = "grey", main = "User rating",
     freq = TRUE, xlim = c(1, 10), xlab = "Rating")
hist(movies$budget, col = "grey", main = "Film budget",
     freq = FALSE, xlab = "Budget")

Box and whisker plots

A box and whisker plot, sometimes called simply a boxplot, is another way to represent continuous data. This kind of plot is particularly useful for comparing two or more groups, by placing the box-plots side-by side. Figure \@ref(fig:moviesbox1) shows box-plots of film length for different categories.

op = par(mfrow = c(2, 1), mar = c(3, 3, 3, 1))
boxplot(sub_movies$length, ylab = "Film length")
boxplot(sub_movies$length ~ sub_movies$mpaa, ylab = "Film length")
par(op)

\indent The central bar in the "box" is the sample median. The top and bottom of the box represent the upper and lower sample quartiles. Just as the median represents the 50% point of the data, the lower and upper quartiles represent the 25% and 75% points respectively.

\indent The lower whisker is drawn from the lower end of the box to the smallest value that is no smaller than 1.5IQR below the lower quartile. Similarly, the upper whisker is drawn from the middle of the upper end of the box to the largest value that is no larger than 1.5IQR above the upper quartile. Points outside the whiskers are classified as outliers.

\indent To draw box and whisker plots in R, we use the boxplot() function. So to construct a boxplot of the movie lengths \@ref(fig:moviesbox1) a, we have

The boxplot() function also accepts formula notation^[Formula notation in R is of the form response_variable ~ explanatory_variable. It is a common construct for specifying model formulae in statistical routines with R but can also be used for Base graphics.]. For example,

## Formula pulling out columns from the data individually
boxplot(sub_movies$length ~ sub_movies$mpaa)
## Formula and data arguments, looks a bit cleaner
boxplot(length ~ mpaa, data = sub_movies)

constructs box and whisker plots for each MPAA rating while

boxplot(length ~ Action + Romance, data = sub_movies)

separates length conditional whether it is a romantic and/or an action movie.

jr-packages/jrIntroduction documentation built on May 24, 2020, 5:19 p.m.