Statistics 744, McMaster University

Final project

Your final project will be done with in a group of size 1-3. The profs will help you pair up (and check with us if you really think you want to go alone).

The project should consist of an exploration of some data. There should be a consistent topic but you needn’t use only a single data file.

Your project should consist roughly of a brief description of the data set, and three or four well-thought-out visualizations (this is a guide, you could have a few as one really fancy one, or as many as ten simple ones that work really well together).

Your presentation will be in person (or at least synchronous), using some sensible format; we encourage reproducible formats but Powerpoint is acceptable. Aim for 10 minutes and be prepared to answer questions.

Your writeup should be in two parts: the main text will be just the text necessary to support the story that you’re trying to tell, and there should be a methods part explaining briefly some of the visualization choices that you made, and if necessary briefly explaining decisions or data cleanup you had to do. Your project should be (very roughly) 10 pages long. Do not include code in your writeup except (if necessary) short snippets illustrating some technical issue you solved.

Please submit your writeup as a PDF or HTML file by noon on 22 December. Submit by pushing to your repository (ideally with all of the scripts used to construct the visualizations from raw data) and letting us know the name of the primary file.

Project milestone (due November 12)

Please submit a brief (0.5-2 pages) prospectus outlining what data you will use and some ideas about the story you wish to tell.

You should post this on a github repo. This can either be a directory in an existing repo, or a new repo to which one of you invites both of us (bbolker, dushoff) and your group members.

Weekly assignments

week 7

1. due Fri Nov 5: read Wilke ch 17-20 and post a discussion question on the Teams channel.

week 6

1. due Fri Oct 29: Make some sort of data visualization using ideas from the last two weeks – that is, something that is either interactive, animated, or explicitly spatial (e.g., map based).

2. due Mon Nov 1 (by 2:30 pm): Write one paragraph about a possible project you might consider doing. You should talk about a possible data set, and also about what kind of visualization interests you. Please post your question on the “project discussion” Teams channel.

week 5

1. due Fri Oct 22: Fit a model to real data, or find a table of model results, and one or two plots illustrating a statistical inference. Let us know if you need help finding a data set, or a plausible scientific question to attack with a model.

2. Read Sections 12-16 of Wilke’s Fundamentals of Data Visualization (and (due Weds Oct 20) post a discussion question on the Teams channel)

week 4

No written assignment. Please read Sections 12-16 of Wilke’s Fundamentals of Data Visualization

for Monday

• e-mail us one discussion question or comment about Wilke sections 8-11

week 4

For Friday:

1. Find a table that you find hard to read and turn it into one, two or three graphs. Explain what features of the data you are trying to draw attention to, and what story you think your figures tell (or fail to tell).

If you have trouble finding a Table you could consider these:

• Table 3 of McCallum et al. 2017
• Table 3 of Dushoff et al. 2005

for Monday

• read Sections 8-11 of Wilke’s Fundamentals of Data Visualization

week 3

For Friday:

• take a look at Jia You’s online graphic about vaccine here (update, a copy of the full version is here

• to get the data, you can use {r eval=FALSE} readr::read_csv("https://mac-theobio.github.io/DataViz/data/vaccine_data_online.csv") or download the data directly from AAAS or get it from the course data page

• write a short statement (a few sentences) that explains what question you think the graphic is trying to answer, or what pattern it’s trying to display
• based on these data, create 2 ggplots that display the data in a different way from the original; use the 2 plots to illustrate tradeoffs between different graphical principles. (If you’re feeling truly uncreative, one of your plots can replicate the original graphical design.) You do not need to worry about the dynamic-graphics aspect or the historical events shown in the original display
• explain (in a few sentences) why your graph or graphs answer the question better than the original graph
• Put your reproducible code in your main repository directory
  • use either hw2.R (with comments as comments or in hw2.txt) or else hw2.rmd for the code
  • put the output graphs (or rmarkdown output) in a directory called HW2

• extra credit: see if you can find information about vaccine coverage over time (i.e., estimates of fraction vaccinated in the US per year) - not just the date of licensing - and incorporate it in your results (we have no idea if this is possible)

week 2

For Monday:

• find an example (on the web, in a research paper, etc etc etc) of either a great (in your opinion) or a terrible (also in your opinion) data visualization.
• write a explanation of a few sentences that explains what is good about it (e.g. interesting, engaging, funny, makes good use of Cleveland principles, makes a point clearly …) or bad about it (is it ugly, bad, or wrong [see Wilke chapter 1]? inaccessible? how/why?)
• put the explanation and a link to the example (you can upload images to your github repo) in the README file in your GH repository, and e-mail us that you’ve done so. (If you run into git trouble you can e-mail us your assignment instead, but putting it all in the repo is better.) (Plain text only, please - in this course we never want .docx (Word) files. PDF is OK, but plain text [including .R files and Markdown files] are always preferred when they work.)

For Friday’s class:

• read Sections 4-7 of Wilke’s Fundamentals of Data Visualization

week 1

For Wednesday’s class:

• make sure you have R (https://cloud.r-project.org/) and RStudio (https://www.rstudio.com/products/rstudio/download/) installed. Please make sure you have at least version 4.0 of R.
• set up a GitHub account (https://github.com) if you don’t already have one (see https://beta.rstudioconnect.com/jennybc/happy-git-with-r/github-acct.html for more information)

For Friday’s class:

• watch Rauser’s “How Humans See Data”
• read Sections 1-3 of Wilke’s Fundamentals of Data Visualization
• read and follow chapters 1, 5, 7, 8, 10 of "“Happy Git with R”"(they’re short!) reading: “Git stuff (; GitHub with GitHub desktop; GitHub on the command line); tidyverse stuff (Software Carpentry lessons on dplyr and tidyr; tidyverse cheat sheets (see especially”“data import”" and "“data transformation”“))”
  • create an (empty except for a README file) repository called Stat744, and add dushoff and bbolker as collaborators
  • email us to let us know you have finished, and to tell us your github username

General

For the first part of the course, there will be a short assignment every week. These will be promulgated usually on Sunday, and you are advised to start before class on Monday. Assignments are technically due Friday at 16:30. This is to encourage you to finish quickly if you can, and move on with your life, not to give you pressure.

If your assignment will be late, please email us by Friday at 16:30 to let us know when you will hand it in. This should be before 16:30 on Monday, unless we give permission for special circumstances.

Submit your assignment by email to macdataviz@gmail.com. Your assignment is not done until you do this.