Predicting Car Prices Based on Certain Characteristics

Authors: Haobo(Harbor) Zhang, Jiaying Liao, Ning Wang, Xiwen Wei

About

In this project, we attempty to build a regression model to predict the car price given several charateristics of the car. The model will be selected using Lasso and Ridge regularizations. We also used 10-fold cross-validation to evaluate the performance of candidate models. The final model was selected by Lasso regularization and it includes idth, curb-weight, and horsepower as its variables.

The data we used was from the The Automobile Data Set that was created by Jeffrey C. Schlimmer in 1985. The data were collected from https://archive.ics.uci.edu/ml/machine-learning-databases/autos/imports-85.data.

Report

The analysis report can be found here.

In pdf and in html.

Usage

We use a Docker container image to make the computational environment for this project reproducible.

First, clone this repo in your terminal by git clone https://github.com/DSCI-310/dsci-310-group-07.git.

Navigate to the root of this project by cd dsci-310-group-07.

Docker

Make sure your working directory contains this Dockerfile.

Then you can obtain the docker image in two different ways:

• Pull from dockerhub (recommended, faster):

  • Go to this webpage.
  • Copy the command on the bottom right, which is

  docker pull wxw1026/dsci-310-group-07:latest
  

  • Paste the command on your terminal and wait for pulling.
  • When the pulling is done, type

  docker images wxw1026/dsci-310-group-07
  

  in your terminal and you will be able to see the image pulled.

  • After obtaining the docker image, you can run this on localhost:8787 by

  docker run -it --rm -p 8787:8787 -e PASSWORD=12345 -v /$(pwd):/home/rstudio/project wxw1026/dsci-310-group-07
  

• Build it locally:

  • Type

  docker build -t dsci-310-group-07 . -f Dockerfile
  

  in your terminal.

  • Wait for installation. It may take minutes.

  • then after obtaining the docker image, you can run this on localhost:8787 by

  docker run -it --rm -p 8787:8787 -e PASSWORD=12345 -v /$(pwd):/home/rstudio/project dsci-310-group-07
  

Open localhost:8787/ on your browser. You will see a login page. The username is rstudio and the password is 12345.

After signing in, you can see the project on /home/rstudio/project.

On the top right panel, open the terminal in rstudio container.

Make

1. In your terimal, run the command make report. It will clean all the previous output (including dataset in .csv and plots in .png) and generate all the new output in need.

2. You can also run make clean to reset your work.

3. Now cd analysis/ and then open report.pdf or open report.html to read the report.

Dependencies

R version 4.1.3 and R packages as follows:

• remotes:2.4.2
• tinytex:0.44
• devtools:2.4.3
• carpriceprediction:1.0.0
• glmnet:4.1-4
• tidyverse:1.3.1
• testthat:3.1.3
• here:1.0.1
• knitr:1.38
• ggcheck:0.0.4
• rmarkdown:2.13
• bookdown:0.26
• cowplot:1.1.1
• docopt:0.7.1

License Information

This project is offered under the Attribution 4.0 International (CC BY 4.0) License. The software provided in this project is offered under the MIT open source license. See the license file for more information.