Porsche Price (2)

library(tidyverse)
library(Stat2Data)

data("PorschePrice")

Using the PorschePrice data, fit a linear model of Price ~ Mileage

lm(Price ~ Mileage, data = PorschePrice)

## 
## Call:
## lm(formula = Price ~ Mileage, data = PorschePrice)
## 
## Coefficients:
## (Intercept)      Mileage  
##     71.0905      -0.5894

Add a variable called y_hat to the PorschePrice dataset with the predicted y values

y_hat <- lm(Price ~ Mileage, data = PorschePrice) %>%
  predict()

PorschePrice <- PorschePrice %>%
  mutate(y_hat = y_hat)

Add another variable called residuals to the PorschePrice dataset with the residuals

PorschePrice <- PorschePrice %>%
  mutate(residuals = Price - y_hat)

Create a plot that examines the “linearity” and “constant variance” assumptions.

ggplot(PorschePrice, aes(x = y_hat, y = residuals)) +
  geom_point() + 
  geom_hline(yintercept = 0) + 
  labs(title = "Residuals vs fits plot",
       x = "predicted Price")

Create a plot that examines the “normality” assumption

ggplot(PorschePrice, aes(sample = residuals)) + 
  geom_qq() + 
  geom_qq_line() + 
  labs(title = "Normal Quantile Plot")

ggplot(PorschePrice, aes(residuals)) + 
  geom_histogram(bins = 15) + 
  labs(title = "Histogram of residuals")