AE 06: Transformations

House Prices

Important

  • Open RStudio and create a subfolder in your AE folder called “AE-06”.

  • Go to the Canvas and locate your AE 06 assignment to get started.

  • Upload the ae-06.qmd files into the folder you just created. The .qmd and HTML responses are due in Canvas. You can check the due date on the Canvas assignment.

library(tidyverse)
library(ggformula)
library(yardstick)
library(mosaic)
library(broom)
library(knitr)
library(patchwork) #arrange plots in a grid

Data

The data set for this assignment contains house sale prices for King County, which includes Seattle, from homes sold between May 2014 and May 2015. It was originally Obtained from Kaggle.com. Our goal will be to try and predict the sale price of a house from the houses square footage.

Variables

  • Outcome
    • price: the sale price in dollars
  • Predictor
    • sqft_living: the square footage of the home

Exercise 0

The dataset that we’ll be using today is called house_prices and is contained in the moderndive package. Load the moderndive package.

Exercise 1

Generate a scatter plot of the data. If you finish before other groups, visualize what a linear model would look like if fitted to this data. Based on what you see, do you think the conditions required of linear regression will be met?

Exercise 2

Generate a scatter plot of the data but take the log of price. You can do this by putting log(price) ~ sqft_living instead of price ~ sqft_living in your call to gf_point. Do you think the data is ready for a linear regression model?

Exercise 3

Generate a scatter plot of the data but take the log of price and a power transformation of sqrt_living. You can do this by putting log(price) ~ sqft_living^(power) instead of price ~ sqft_living in your call to gf_point. Once you’re done, your data should be ready for modeling.

Exercise 4

Fit the model above by using the lm function. Output the coefficients of your model. Can do this by using the formula (the thing with the ~), just put I() around any power. For example I(sqrt_living^(1/2)) instead of sqrt_living^(1/2).

Exercise 5 (Difficult)

Write your final model in the board in two ways. One with \(\log(Y)\) and one with \(Y\) is on the right side of the equal sign. You will want to use the properties of logs.

Exercise 6 (Time Permitting)

Analyze the residuals for the model you created model.

To submit the AE:

Important
  • Render the document to produce the HTML file with all of your work from today’s class.
  • Upload your QMD and HTM files to the Canvas assignment.