Lec01: Welcome!

# Lec01: Welcome!
## Stat41: Data Viz
### Prof Amanda Luby
### Swarthmore College
### 2021/01/04 (updated: 2021-01-04)

---

# Today:

## 1. Overview of the class

## 2. Basics of Data Viz

## 3. Meet your group!

---
class: inverse, center, middle

# Overview of the class

---

# Syllabus

Preliminary/visual version  and full "rules and regulations" version available [on the course website](https://aluby.domains.swarthmore.edu/stat041/syllabus.html).

Please read **both** in detail at some point this week

We'll only hit the highlights today

Please interrupt me with any questions!

---

# Three main pieces of this course:

**Lecture/Discussion:**

+ M-Th 11a-12p ET
+ synchronous (exceptions for extreme time zone differences)
+ attendance expected

**Labs:**

+ M-Th 1p-2p ET
+ asynchronous option
+ I'll be on Slack (and maybe zoom)

**Projects:**

+ Weekly mini-projects
+ Final project

Everything will be posted on the [course website](https://aluby.domains.swarthmore.edu/stat041), except for a few readings which will be posted on our private drive, and you will turn things in via moodle.

---

# Lecture/Discussion

It's *extremely* easy to fall behind in these intense class experiences. I think it's important to have regular meeting time together to check in, stay on track, and get help if needed.

1. Talk to other humans we don't live with
2. Synthesize readings
3. Share cool stuff with each other
4. Apply concepts + discuss in small groups
5. Ask questions about the lab
]

Your responsibilities:

**Before "class"**: Short reading or video, prepare any Q's

**During "class"**: Ask Q's, share visualizations you've come across, keep track of discussion in Jamboards

**After "class"**: Do the lab, keep track of questions

]

---

# Labs

The best way to learn is by doing. The daily labs will give you guided practice making graphs in R, and most will include an open-ended portion to challenge yourself as much or as little as you like.

If you want to build a large portfolio, the labs are a great place to do so!

While we'll have a lab each class day, they are not "due" until Sunday, so if you run out of steam it is perfectly OK to finish them later.

The labs will often involve replicating existing graphs or tweaking them in some way. My goal for this class is that you all learn skills for making graphs *after* this class, which often involves searching for and reading others' work. This is a hard skill to learn, so we're going to practice it!

**A quick note on time commitment:** My expectation is that you will spend 3-4 hours on class days, plus 6-10 hours on projects per week, so somewhere in the 20-25 hours per week range. If you're spending much more or less time than that, please let me know.

---

# Projects
.pull-left[
### Weekly Mini-Projects

(1) Replicating a published graph + Improving it + Ruining it

(2) "Blog post"-style walk-through of fitting a model

(3) Explanation of a statistical concept from a previous class using interactivity
]

Build data visualizations of a dataset of your choice! Will include *both* a collection of visualizations and a 10-ish page paper. Each Sunday, you'll have a "milestone" due:

1. Find topic + data, making sure you can load it into R
2. EDA + introduction written
3. Rough Draft (minus interactivity)

More details to come on Thursday.

**Please reserve Thurs, Jan 28 11a-2:30p ET for presentations/demos**
]

---

# Contacting me

The best way to get your questions answered is to reach out to me via **slack** (you should have received an invitation this morning):

+ R Help
+ Clarifications about labs or projects
+ Logistical questions
+ Setting up a meeting

If you have questions about the course material, there's a good chance others do too!

Sensitive/private topics, or anything you want a more formal record of, can still come to my email (it just might take me longer to respond)

---

### Other guidelines:

We're *still* in a pandemic

Almost all of you are learning from home

You are all in different places when it comes to bandwidth you can commit to this class

### **THAT'S OKAY**

We're going to normalize **not apologizing** for things that aren't our fault:

- Computer/Internet/Server problems

- Sick family members

- Helping kids with virtual school

Instead, we'll say *thank you for understanding*.

I will sometimes get over-excited about what we are doing in this course - **please don't mistake this excitement for pressure or expectation.**

---

# Questions?

---
class: inverse, center, middle

# Basics of Data Viz

---

# Let's "look" at some data:

![](images/Tufte_Anscombe0.png)

---
# This is called "Anscombe's Quartet":

![](images/Tufte_Anscombe1.png)

All four datasets have the **exact same** linear model (minus the residuals).

Source: Tufte's *The Visual Display of Quantitative Information* (a Data Viz classic)

---

![](images/Tufte_Anscombe2.png)
--

+ **Always graph your data before you calculate a statistic**
+ **Understand the impact of outliers on statistical properties**
+ **Computers are dumb!**
---

# More gems from Tufte

* induce viewer to think about substance, not graphical methodology

* avoid **distorting** the data or letting **decoration** get in the way

* make large, complicated datasets more coherent

* encourage comparison of different pieces of data

* reveal data at several levels of detail

* describe, explore, tabulate, identify relationships

* be closely integrated with statistical/verbal descriptions

* use **consistent graph design**

**Avoid graphs that lead viewers to make misleading conclusions**

---

# Example 1

<center> 
![](images/FL-gun-deaths.png)

---
# Example 2

![](images/GA-DPH-non-chron-cases.jpeg)
Source: GA Department of Health
---
# Example 3

![](images/nature-symptom-venn.jpeg)
---
# Example 4

![](images/covid-tornado-plot.jpeg)
---
# Example 5

real graphic from a talk  
([`crabs` dataset](https://stat.ethz.ch/R-manual/R-devel/library/MASS/html/crabs.html) in `MASS` package)

![](Lec01_files/figure-html/unnamed-chunk-2-1.png)

---

# Meet your group!

---

# Your to-do list:

### 1. Introductions: name, year, favorite class (so far 😉)
### 2. Class expectations (first page on jamboard)
### 3. Talk through the first 4 examples using the *Five Qualities of Great Visualizations* and Tufte's guidelines on your  [jamboard](https://drive.google.com/drive/folders/10Qc2mvHq8SHw2JsXWAUU5a0u-9X0YAmZ?usp=sharing)

---

# Looking forward:

* [Lab 01](https://aluby.domains.swarthmore.edu/stat041/labs.html) this afternoon, due on moodle on Sunday
  - Come back here at 1pm ET to work synchronously with your group 
  - Ask Q's on Slack

* Tomorrow: 
  - Historical data viz 
  - Grammar of Graphics
  
--

* [Project 1 prompt](https://aluby.domains.swarthmore.edu/stat041/projects.html) on course website