Plots/ggplot2

The ggplot2 library is a part of the tidyverse package.
Citing the original description:

“ggplot2 is a system for declaratively creating graphics, based on The Grammar of Graphics. You provide the data, tell ggplot2 how to map variables to aesthetics, what graphical primitives to use, and it takes care of the details”.

Remember to use library( tidyverse ) at the beginning of your R Markdown document.
Load the pulse.csv table into the pulse variable.
Practice in an R Markdown document and Knit it regularly to see the generated report.

A simple scatter plot

When you use ggplot2, you compose a plot by “adding” components. Here are some examples:

ggplot( pulse ): Always the first component. It starts a new plot for data from the variable pulse (a tibble or a data.frame).
aes( x = weight ): aesthetics – Sets usage of the data column weight for the x axis.
geom_point(): Sets the scatter plot geometry (requires at least x and y).

Now enter the following code and create a simple scatter plot (also Knit the document):

ggplot( pulse ) +
  aes( x = weight, y = height ) +
  geom_point()

Aesthetics

Open Help for the command geom_point and find there the section Aesthetics.
There you can find keywords to specify aesthetics known by geom_point: position, color, shape of the points, …

Fixed: all points identical

Type the following example code and generate the plot.
Regenerate the plot for another color (google-search to find allowed “R color names”):

ggplot( pulse ) +
  aes( x = weight, y = height ) +
  geom_point( color = "darkblue" )

Type the following code with different values of size (e.g.: 0.5, 1, 2, 5) to see the effect:

ggplot( pulse ) +
  aes( x = weight, y = height ) +
  geom_point( color = "darkblue", size = 5 )

Now, add and modify alpha (range from 0.0 to 1.0) to change transparency:

ggplot( pulse ) +
  aes( x = weight, y = height ) +
  geom_point( color = "darkblue", size = 5, alpha = 0.5 )

Next, google-search to find allowed “R point shapes”.
Shapes are numbered (range: 0..25). Try:

ggplot( pulse ) +
  aes( x = weight, y = height ) +
  geom_point( shape = 24, size = 3 )

Note, that some shapes (21..25) might be drawn with two different colors.
They are specified with color and fill aesthetics. Try:

ggplot( pulse ) +
  aes( x = weight, y = height ) +
  geom_point( shape = 24, size = 3, color = "blue", fill = "lightblue" )

Variable: points diffier

Variable aesthetics need to be specified as arguments to aes( ... ) function.
For example, type the following code to make color of the points dependent on the exercise column of pulse.
Note, that the library recognizes that the variable is categorical and creates an appropriate legend.
Scale colors are assigned automatically (other possibilities will be discussed in the next section).

ggplot( pulse ) +
  aes( x = weight, y = height, color = exercise ) +
  geom_point()

The pulse1 variable is continuous (not categorical).
Try the following and observe the change in the legend (as above, the scale is created automatically).

ggplot( pulse ) +
  aes( x = weight, y = height, color = pulse1 ) +
  geom_point()

Combined

Naturally, it is possible to combine variable and fixed aesthetics. Try:

ggplot( pulse ) +
  aes( x = weight, y = height, size = pulse1, fill = exercise ) +
  geom_point( shape = 21, alpha = 0.65, color = "black" )

Note, that the notation is quite flexible.
Check, that indeed the same chart is generated by:

ggplot( pulse, aes( x = weight ) ) +
  aes( y = height ) +
  aes( fill = exercise ) +
  geom_point( shape = 21, alpha = 0.65, color = "black", aes( size = pulse1 ) )

↓ Practice ← Groups Plot axes/scales →