Schedule

Last time

Regession

Today

Plotting in R (especially ggplot2)

Plotting

The default option: plot()

While we'll quickly move on to other options, R's plot() function (in the default graphics package) is a great tool for basic data exploration—it's fast, simple, and flexible.

Plotting

The default option: plot()

While we'll quickly move on to other options, R's plot() function (in the default graphics package) is a great tool for basic data exploration—it's fast, simple, and flexible.

In fact, plot() is a generic function, that works for many classes.

Plotting

The default option: plot()

While we'll quickly move on to other options, R's plot() function (in the default graphics package) is a great tool for basic data exploration—it's fast, simple, and flexible.

In fact, plot() is a generic function, that works for many classes.

General arguments for plot():

• x and y for coordinates
• type = {"p"oints, "l"ines, etc.} (optional)
• xlab, ylab, main, and sub for axis labels and (sub)title (optional)
• col and pch for color and plot character (optional)
• lty and lwd for line type, and line width (optional)

ggplot2

The grammar

The ggplot2 package offers an incredibly flexible, diverse, and powerful set of functions for creating graphics in R.

ggplot2

The grammar

The ggplot2 package offers an incredibly flexible, diverse, and powerful set of functions for creating graphics in R.

The gg stands for the grammar of graphics.

ggplot2

The grammar

The ggplot2 package offers an incredibly flexible, diverse, and powerful set of functions for creating graphics in R.

The gg stands for the grammar of graphics.

ggplot2

1. centers on a data frame (the data argument)
2. maps variables to aesthetics (the aes argument)
3. layers geometries to build up your graphic

Note The package is called ggplot2, but the main function is ggplot().

ggplot2

ggplot()

Main arguments

1. data Your dataset. As a data frame (or tibble).

ggplot2

ggplot()

Main arguments

1. data Your dataset. As a data frame (or tibble).

2. aes() Maps variables in data to "aesthetics" like x, color, shape.

ggplot2

ggplot()

Main arguments

1. data Your dataset. As a data frame (or tibble).

2. aes() Maps variables in data to "aesthetics" like x, color, shape.

Example A time series of problems, color defined by money

library(ggplot2)
ggplot(
  data = pretend_df,
  aes(x = time, y = problems, color = money)
)

ggplot2

Layers

The ggplot() function doesn't plot anything—it sets up the plot.

To create the actual figure, you layer geometries (e.g., geom_point()),

ggplot2

Layers

The ggplot() function doesn't plot anything—it sets up the plot.

To create the actual figure, you layer geometries (e.g., geom_point()),
scales (e.g., scale_color_manual()),

ggplot2

Layers

The ggplot() function doesn't plot anything—it sets up the plot.

To create the actual figure, you layer geometries (e.g., geom_point()),
scales (e.g., scale_color_manual()), and other options (e.g., xlab()).

ggplot2

Layers

The ggplot() function doesn't plot anything—it sets up the plot.

To create the actual figure, you layer geometries (e.g., geom_point()),
scales (e.g., scale_color_manual()), and other options (e.g., xlab()).

You add layers using the addition sign (+).

ggplot2

Layers

The ggplot() function doesn't plot anything—it sets up the plot.

To create the actual figure, you layer geometries (e.g., geom_point()),
scales (e.g., scale_color_manual()), and other options (e.g., xlab()).

You add layers using the addition sign (+).

Example A time series of problems, color defined by money

library(ggplot2)
ggplot(
  data = pretend_df,
  aes(x = time, y = problems, color = money)
) +
geom_point() + geom_line()

ggplot2

Saving plots

You can save your ggplot2-based figures using ggsave().

ggplot2

ggsave() Option 1

By default, ggsave() saves the last plot printed to the screen.

# Create a simple scatter plot
ggplot(data = fun_df, aes(x = x, y = y)) +
geom_point()
# Save our simple scatter plot
ggsave(filename = "simple_scatter.pdf")

ggplot2

ggsave() Option 1

By default, ggsave() saves the last plot printed to the screen.

# Create a simple scatter plot
ggplot(data = fun_df, aes(x = x, y = y)) +
geom_point()
# Save our simple scatter plot
ggsave(filename = "simple_scatter.pdf")

Notes

• This example creates a PDF. Change to ".png" for PNG, etc.
• There several helpful, optional arguments: path, width, height, dpi.

ggplot2

ggsave() Option 2

You can assign your ggplot() objects to memory

# Create a simple scatter plot named 'gg_points'
gg_points <- ggplot(data = fun_df, aes(x = x, y = y)) +
geom_point()

ggplot2

ggsave() Option 2

You can assign your ggplot() objects to memory

# Create a simple scatter plot named 'gg_points'
gg_points <- ggplot(data = fun_df, aes(x = x, y = y)) +
geom_point()

We can then save this figure with the name gg_points using ggsave()

# Save our simple scatter plot name 'ggsave'
ggsave(
  filename = "simple_scatter.pdf",
  plot = gg_points
)

Resources

There's always more

ggplot2

1. RStudio's cheat sheet for ggplot2.
2. ggplot2 reference index
3. The tidyverse page on ggplot2.
4. Hadley Wickham's on Data visualization in his data science book.

Table of contents