ae-3-duke-forest-notes.qmd

---
title: "AE 3: Duke Forest houses"
subtitle: "Checking model conditions"
author: "Add your name here"
format: pdf
editor: visual
---

## Packages

```{r load-packages}
#| message: false
library(tidyverse)
library(tidymodels)
library(openintro)
library(knitr)
```

## Predict sale price from area

```{r}
df_fit <- linear_reg() %>%
  set_engine("lm") %>%
  fit(price ~ area, data = duke_forest)

tidy(df_fit) %>%
  kable(digits = 2)
```

\pagebreak

## Model conditions

### Exercise 1

The following code produces the residuals vs. fitted values plot for this model.
Comment out the layer that defines the y-axis limits and re-create the plot.
How does the plot change?
Why might we want to define the limits explicitly?

```{r}
df_aug <- augment(df_fit$fit)

ggplot(df_aug, aes(x = .fitted, y = .resid)) +
  geom_point() +
  geom_hline(yintercept = 0, linetype = "dashed") +
  ylim(-1000000, 1000000) +
  labs(
    x = "Fitted value", y = "Residual",
    title = "Residuals vs. fitted values"
  )
```

A symmetric axis allows us to evaluate more easily whether the residuals are randomly scattered around the $y = 0$ line.

\pagebreak

### Exercise 2

Improve how the values on the axes of the plot are displayed by modifying the code below.

```{r}
ggplot(df_aug, aes(x = .fitted, y = .resid)) +
  geom_point() +
  geom_hline(yintercept = 0, linetype = "dashed") +
  labs(
    x = "Fitted value", y = "Residual",
    title = "Residuals vs. fitted values"
    ) +
  scale_x_continuous(labels = label_dollar()) +
  scale_y_continuous(labels = label_dollar(), limits = c(-1000000, 1000000)) 
``