Data visualization

class: center, middle, inverse, title-slide

# Data visualization
### Yue Jiang
### Duke University

---

### A disclaimer

The following material was used during a live lecture. Without the accompanying
oral comments and discussion, the text is incomplete as a record of the
presentation. A full recording may be found via Zoom on the course Sakai site.

---

### Why visualization?

.small[Image source: Adapted from Wang, 2020, re: Kobak and Berens (Nat. Commun. 2019) and Cao et al. (Nature 2019).]

---

### Why visualization?

.small[Image source: Playfair. 1786 trade balance chart, Commercial and Political Atlas.]

---

### Why visualization?

```r
dat
```

```
##    x1 x2 x3 x4    y1   y2    y3    y4
## 1  10 10 10  8  8.04 9.14  7.46  6.58
## 2   8  8  8  8  6.95 8.14  6.77  5.76
## 3  13 13 13  8  7.58 8.74 12.74  7.71
## 4   9  9  9  8  8.81 8.77  7.11  8.84
## 5  11 11 11  8  8.33 9.26  7.81  8.47
## 6  14 14 14  8  9.96 8.10  8.84  7.04
## 7   6  6  6  8  7.24 6.13  6.08  5.25
## 8   4  4  4 19  4.26 3.10  5.39 12.50
## 9  12 12 12  8 10.84 9.13  8.15  5.56
## 10  7  7  7  8  4.82 7.26  6.42  7.91
## 11  5  5  5  8  5.68 4.74  5.73  6.89
```

---

### Why visualization?

```r
dat %>% 
  summarize(meanx1 = mean(x1),
            meanx2 = mean(x2),
            meanx3 = mean(x3),
            meanx4 = mean(x4))
```

```
##   meanx1 meanx2 meanx3 meanx4
## 1      9      9      9      9
```

```r
dat %>% 
  summarize(meany1 = mean(y1),
            meany2 = mean(y2),
            meany3 = mean(y3),
            meany4 = mean(y4))
```

```
##   meany1 meany2 meany3 meany4
## 1    7.5    7.5    7.5    7.5
```

---

### Why visualization?

```r
dat %>% 
  summarize(sdx1 = var(x1),
            sdx2 = var(x2),
            sdx3 = var(x3),
            sdx4 = var(x4))
```

```
##   sdx1 sdx2 sdx3 sdx4
## 1   11   11   11   11
```

```r
dat %>% 
  summarize(sdy1 = sd(y1),
            sdy2 = sd(y2),
            sdy3 = sd(y3),
            sdy4 = sd(y4))
```

```
##   sdy1 sdy2 sdy3 sdy4
## 1 2.03 2.03 2.03 2.03
```

---

### Why visualization?

```r
dat %>% 
  summarize(cor1 = cor(x1, y1),
            cor2 = cor(x2, y2),
            cor3 = cor(x3, y3),
            cor4 = cor(x4, y4))
```

```
##    cor1  cor2  cor3  cor4
## 1 0.816 0.816 0.816 0.817
```

---

### Why visualization?

```r
lm(dat$y1 ~ dat$x1)$coef
```

```
## (Intercept)      dat$x1 
##         3.0         0.5
```

```r
lm(dat$y2 ~ dat$x2)$coef
```

```
## (Intercept)      dat$x2 
##         3.0         0.5
```

```r
lm(dat$y3 ~ dat$x3)$coef
```

```
## (Intercept)      dat$x3 
##         3.0         0.5
```

```r
lm(dat$y4 ~ dat$x4)$coef
```

```
## (Intercept)      dat$x4 
##         3.0         0.5
```

---

### Why visualization?

```r
par(mfrow = c(2,2))
plot(dat$x1, dat$y1, xlab = "X1", ylab = "Y1", pch = 19)
plot(dat$x2, dat$y2, xlab = "X2", ylab = "Y2", pch = 19)
plot(dat$x3, dat$y3, xlab = "X3", ylab = "Y3", pch = 19)
plot(dat$x4, dat$y4, xlab = "X4", ylab = "Y4", pch = 19)
```

---

### Why visualization?

---

### What can go wrong?

.small[Image source: CDC, 2020]

---

### What can go wrong?

.small[Image source: Andrew Yang campaign, 2020]

---

### What can go wrong?

---

### What can go wrong?

---

### What can go wrong?

.small[Image source: Leonhardt 2020, CNBC]

---

### What can go wrong?

.small[Image source: Potter et al., Fundamentals of Nursing]

---

### What can go wrong?

.small[Image source: Waugh and Clements, AQA-level History: Making of a Superpower: US 1865-1975]

---

### What can go wrong?

.small[Image source: He and Williams for Bloomberg, 2020]

---

### What can go wrong?

.small[Image source: Gadarian, Goodman, and Pepinsky (Draft figure from SSRN 2020)]

---

### What can go wrong?

.small[Image source: Gadarian, Goodman, and Pepinsky (SSRN 2020)]

---

### Some best practices

- Tell a story; have one in mind when deciding to use a visualization
- Every figure should serve a clear purpose
- A figure should stand on its own 
- Keep it simple and clear

---

### Some best practices

Make sure the type of visualization chosen is correct given the number of
variables included, their types, and their relationships. Some helpful links
are below:

- [The Data Visualization Catalogue](https://datavizcatalogue.com/)
- [The Financial Times Visual Vocabulary](https://github.com/ft-interactive/chart-doctor/tree/master/visual-vocabulary)

---

### Some best practices

Often, journals will present their own guidelines for visualizations. For 
instance, to submit an article to JAMA, a style guide for figures and their
purposes is provided [here](https://jamanetwork.com/journals/jama/pages/instructions-for-authors#SecTableofFigureRequirements) and contains the most often used 
plots for this particular journal.

---

### Workshop