PADP 7120 Data Applications in PA

class: center, middle, inverse, title-slide

.title[
# PADP 7120 Data Applications in PA
]
.subtitle[
## Data Visualization
]
.author[
### Alex Combs
]
.institute[
### UGA | SPIA | PADP
]
.date[
### Last updated: February 03, 2023
]

---

# Outline

- Choosing the appropriate graphs

- Guidelines of good visualization

---
# Some Data

|major                                                             | grad_median|
|:-----------------------------------------------------------------|-----------:|
|Electrical Engineering Technology                                 |       85000|
|Anthropology And Archeology                                       |       65000|
|Materials Engineering And Materials Science                       |       92000|
|Electrical, Mechanical, And Precision Technologies And Production |       62000|
|Library Science                                                   |       52000|

- Includes 173 graduate degree majors

---
# Choosing appropriate graph

- Which graph might we use if we want to visualize the distribution of median pay?

---
# Histogram

.pull-left[
![](Visualizations_files/figure-html/unnamed-chunk-3-1.png)
]

.pull-right[

```r
ggplot(college_grad_students, 
       aes(x = grad_median)) +
  geom_histogram(bins = 30)
```
]

---
# Exploratory vs. explanatory

.center[
![](lectures_files/twogoals.png)
]

---
# An exploratory graph

.pull-left[
![](Visualizations_files/figure-html/unnamed-chunk-6-1.png)
]

.pull-right[

```r
ggplot(gm_americas, aes(x = gdpPercap, y = lifeExp)) +
  geom_point() +
  geom_smooth(method = "lm", se = FALSE)
```
]
---
# An explanatory graph
.pull-left[
![](Visualizations_files/figure-html/unnamed-chunk-8-1.png)
]
.pull-right[

```r
ggplot(gm_americas, aes(x = gdpPercap, y = lifeExp)) +
  geom_point() +
  geom_smooth(se = FALSE, linetype = "dashed", color = "gray", span = 1) +
  scale_x_continuous(labels = scales::dollar) +
  labs(title = "Life Expectancy and Wealth in the Americas",
       subtitle = "2007",
       y = "Life Expectancy",
       x = "GDP Per Capita",
       caption = "Source: Gapminder Dataset") +
  theme_classic() +
  theme(text = element_text(size = 14))
```
]

---

# Another graph for an audience

.center[
![](labs_files/lifeExp.png)
]

---
# Qualities of good viz

- Proper graph given the data

- Display an accessible complexity of detail

- Don't distort perception

- Avoid content-free decoration--chart junk

- Maximize data-to-ink ratio

---
# Chart junk?

.center[
![:scale 80%](lectures_files/monstrouscosts.png)
]

---
# Distortion

- Reproduction of a graph in the *New York Times* about a crisis in democracy

.center[
![](lectures_files/badcrisis.png)
]

---
# Distortion

.center[
![:scale 85%](lectures_files/bettercrisis.png)
]

- Previous graph focused on the percentage of people who gave a rating of 10

---
# Distortion

.center[
![](lectures_files/3dgraph.png)
]

---
# Distortion

- Law school enrollment trend

.center[
![](lectures_files/lawaxes.png)
]

- Consider how axes change the perception of scale.

---
# Possible distortion

.center[
![](lectures_files/showzero.png)
]

- Depending on context, bar charts can lead people to think values inside the bars are more likely than values above.

- More examples of bad graphs can be found [here](https://badvisualisations.tumblr.com)

---
class: inverse, middle, center

# Human perception has strengths and weaknesses

---
# Decoding numerical visualization

.center[
![](lectures_files/encode1.png)
]

- Why simple (i.e., one or two variables) bar graphs or scatterplots are typically better.

---
# Example

.center[
![:scale 70%](lectures_files/badstackbars.png)
]

- Easily compare totals and bottom category, but other three are difficult to compare across bars.

- Also difficult to compare within bars.

- Dodged bar chart would be better.

---
# Example

```
## Warning: ggrepel: 1 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
```

---
# Example

```
## Warning: ggrepel: 1 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
```

---
# Decoding categorical visualization

.center[
![:scale 70%](lectures_files/encode2.png)
]

---
# Find the blue circle

.center[
![:scale 40%](lectures_files/bluecirc1.png)
]
---
# Find the blue circle
.center[
![:scale 40%](lectures_files/bluecirc2.png)
]
---
# Find the blue circle
.center[
![:scale 40%](lectures_files/bluecirc3.png)
]
---
# Find the blue circle
.center[
![:scale 40%](lectures_files/bluecirc4.png)
]
---
# Find the blue circle
.center[
![:scale 40%](lectures_files/bluecirc5.png)
]

---
class: inverse, middle, center

# Let's consider a few graphs you all made in R Chapter 5

---
# Number of counties by state

![](Visualizations_files/figure-html/unnamed-chunk-13-1.png)

---
# Number of counties by state

![](Visualizations_files/figure-html/unnamed-chunk-14-1.png)

---
# Number of counties by state

![](Visualizations_files/figure-html/unnamed-chunk-15-1.png)

---
# Scatterplot

![](Visualizations_files/figure-html/unnamed-chunk-16-1.png)

---
# Scatterplot

![](Visualizations_files/figure-html/unnamed-chunk-17-1.png)

---
# Scatterplot

![](Visualizations_files/figure-html/unnamed-chunk-18-1.png)

---
# Scatterplot

![](Visualizations_files/figure-html/unnamed-chunk-19-1.png)

---
# Scatterplot

![](Visualizations_files/figure-html/unnamed-chunk-20-1.png)

---
# Scatterplot

![](Visualizations_files/figure-html/unnamed-chunk-21-1.png)