10 Smoother

Let’s continue with the 2001 first quarter dataset and add a smoother.

ggplot(housing2001q1, aes(x = Land.Value, y = Structure.Cost)) + 
  geom_point() + 
  scale_x_log10() +
  stat_smooth()

## `geom_smooth()` using method = 'loess'

We can fit a linear model to our dataset

ggplot(housing2001q1, aes(x = Land.Value, y = Structure.Cost)) + 
  geom_point() +
  scale_x_log10() +
  stat_smooth(method = "lm")

We can also specify the formula for the model

ggplot(housing2001q1, aes(x = Land.Value, y = Structure.Cost)) + 
  geom_point() +
  scale_x_log10() +
  stat_smooth(method = "lm", formula = y ~ log(x))

We can turn the turn off the confidence interval

ggplot(housing2001q1, aes(x = Land.Value, y = Structure.Cost)) + 
  geom_point() +
  scale_x_log10() +
  stat_smooth(method = "lm", formula = y ~ log(x), se = FALSE)

Formula is specific to the type of model used. Here we’re using a General Additive Model (GAM).

ggplot(housing2001q1, aes(x = Land.Value, y = Structure.Cost)) + 
  geom_point() +
  scale_x_log10() +
  stat_smooth(method = "gam", formula = y ~ s(x,k=10))