7 Geoms and Statistics

Geometric objects (geoms) define the basic shape of the elements on the plot.

• Every geom has a default statistic
• Every statistic has a default geom

You can get a list of all geoms using the online help in RStudio

help.search("geom_", package = "ggplot2")

Change the size of each bin:

ggplot(housing, aes(x = Home.Value)) +
  geom_histogram(binwidth = 1000)

Add a mapping for the fill color:

ggplot(housing, aes(x = Home.Value, fill = Region)) +
  geom_histogram(binwidth = 1000)

Mapping can also be specified in the geom:

ggplot(housing, aes(x = Home.Value)) +
  geom_histogram(aes(fill = Region), binwidth = 1000)

Same plot can also be created using stat_bin transformation. The default geom for stat_bin is “area”

ggplot(housing, aes(x = Home.Value)) +
  stat_bin(binwidth = 1000)

Change the default geom to "point"

ggplot(housing, aes(x = Home.Value)) +
  stat_bin(geom = "point", binwidth = 1000)

Change the default geom to "line"

ggplot(housing, aes(x = Home.Value)) +
  stat_bin(geom = "line", binwidth = 1000)

7.1 Exercise

Create a subset of housing data from New York since 2000

newyork2k <- subset(newyork, Year >= 2000)

1. Create a plot that includes multiple geometric objects, for example, lines and points.

   

2. Change the shape to be hollow diamond

   HINT: Take a look at Shape Scales in the Data Visualization with ggplot2 Cheat Sheet

   

3. Change the size of the point based on Land.Value and color based on Structure.Cost