Data Visualization (reference card)

Basic ggplot

ggplot() creates blank objects and defines default values
data sets the dataframe
mapping sets what columns of the data frame correspond to different parts of the plot
The mapping argument must be an “aesthetic” create using the aes() function
Add a geom_ to represent the data, e.g., geom_point() for point

ggplot(data = df, mapping = aes(x = col_name_1, y = col_name_2)) +
  geom_point()

Customizing geom

Add optional arguments to the geom to change how it looks

ggplot(data = df, mapping = aes(x = col_name_1, y = col_name_2)) +
  geom_point(size = 3, color = "blue", alpha = 0.5)

Rescaling axes

You can change the scales of things using scale_ functions
To scale axes logarithmically

ggplot(data = df, mapping = aes(x = col_name_1, y = col_name_2)) +
  scale_y_log10() +
  scale_x_log10()

Grouping

Color

Color points depending on the value of a categorical variable

ggplot(df, aes(x = col_name_1, y = col_name_2, color = TREATMENT)) +
  geom_point()

Facets (i.e., subplots)

Split the data into facets/subplots depending on the value of a categorical variable

ggplot(df, aes(x = col_name_1, y = col_name_2)) +
  geom_point() +
  facet_wrap(vars(TREATMENT))

Bar plot

ggplot(df, aes(x = col_name)) +
  geom_bar()

Histograms

ggplot(df, aes(x = col_name)) +
  geom_histogram(fill = "red", bins = 15)

Stacked

ggplot(df, aes(x = col_name_1, color = col_name_2)) +
  geom_histogram()

Unstacked

ggplot(df, aes(x = col_name_1, color = col_name_2)) +
  geom_histogram(position = "identity", alpha = 0.5)

Multiple layers

Add more than one geom to get more than one layer of data
E.g., one layer of point data and one layer showing a smooth of the data

ggplot(df, aes(x = col_name_1, y = col_name_2)) +
  geom_point() +
  geom_smooth(method = "lm")

Changing values across layers

If you set an aesthetic in the geom, it will only apply to that layer
So to color the points by col_name_3, but create one smooth for all points

ggplot(df, aes(x = col_name_1, y = col_name_2)) +
  geom_point(mapping = aes(color = col_name_3)) +
  geom_smooth(method = "lm")

Saving plots

ggsave("name_for_file.jpg")