Plot types

Choosing the right type of plot, depends on your data. Wheather you have continous data, a factor, grouped variables, will define what type of plots you can do. Here, we present a number of common plot types. You can find the code to plot these figures below. Here is a guide that can help you choose the plot type (if we still want to keep it…)

In ggplot, the geom_xxx() funciton defines the type of plot that is drawn. in the example below, we use geom_point(), which makes dots.

Continous variables

Continous data is often visualized using dots, boxplots or bars. A heat map can also be used.

Code

Points

Here is the code to make a dotplot using geom_point():

ggplot(data = iris, aes(x = Sepal.Length, y = Petal.Length, colour = Species, shape = Species)) +
  geom_point(size = 2, alpha = 0.6) +
  labs(title = "Iris data", x = "Sepal Length", y = "Petal Length") +
  scale_color_viridis_d() +
  theme_minimal()

Box

Here is the code to make a boxplot using geom_boxplot():

ggplot(data = iris, aes(x = Species, y = Petal.Length, fill = Species)) +
  geom_boxplot() +
  labs(title = "Iris data", x = "Sepal Length", y = "Petal Length") +
  scale_fill_viridis_d() +
  theme_minimal()

Bars

Here is the code to make a barplot using geom_bar():

ggplot(data = iris, aes(y = Petal.Length, x = Species, fill = Species)) +
  geom_bar(stat = "identity") +
  labs(x = "", y = "") +
  scale_fill_viridis_d() +
  theme_minimal()

Categorical data, factors and groups

To show categorical data, factors and groups we often use boxplots, or other elements. We can use colour, size, shape or linetype to distinguish different groups. Another option is to divide the plot into subplots using the facet function.

Distributions

Histograms and density plots are most commonly used to plot densities. To visualize several density at the same time, boxplots, violin and sina plot are useful.

Boxplots, show the median, quantiles, outliers = useful. Violin plots… Sina even show the data. Ridgeline plots

Proportions

Proportions are often displayed using barplots or stacked barplots.

x~y relationships

To show relationships between two continous variables, we often use scatterplots. Contour lines Correlations between variables can be shown using a correlogram.

Spatial and temporal data

Spatial data is often shown on maps. Temporal data, is often shown with lines/smooth

Uncertainty

Errorbars are often used to show uncertainty of the data or likely values - posterior distribbution/eye plot - confidence bands