Animated and Interactive Visualizations

class: middle, center, title-slide

.title[
# Animated and Interactive Visualizations
]
.subtitle[
## Data Visualization
]
.author[
### Johan Larsson
]
.author[
### Behnaz Pirzamanbein
]
.institute[
### The Department of Statistics, Lund University
]

---

## Animated Visualizations

Animated Visualization
- offers an extra aesthetic that we can map variables to.

- is very useful in **presentations**.

Animating works best when mapping to time or similar process.

### gganimate

.pull-left-60[
[gganimate](https://gganimate.com/) is a ggplot-based solution for producing
animations in R.

You may need to download and install a
**renderer**; we recommend [gifski](https://gif.ski/).
]

.pull-right-40[
<img src="images/gganimate.svg" width="50%" style="display: block; margin: auto;" />

]

---

### Example: Housing Sales in Texas

Start with some data.

```r
# aggregate over years
tx <- select(txhousing, city, year, sales:inventory) %>%
  drop_na() %>%
  group_by(city, year) %>%
  summarize_all(mean, na.rm = TRUE) 
```

.pull-left[
Then build a basic ggplot.

```r
p <- tx %>%
  ggplot(aes(median, 
             listings)) +
    geom_point(alpha = 0.5)
p
```
]

.pull-right[
<div class="figure" style="text-align: center">
<img src="animated-and-interactive-visualizations_files/figure-html/unnamed-chunk-3-1.png" alt="Median sales price and numbers of listings." width="100%" />
<p class="caption">Median sales price and numbers of listings.</p>
</div>
]

---

### Animating the Plot

Adding `ease_aes()` and `transition_*()` to the plot sets up the animation.

Calling the object invokes the `animate()` function,
which does all the heavy-lifting.

.pull-left[

```r
library(gganimate)

anim <- p +
* transition_time(year) +
* ease_aes()

anim # same as animate(anim)
```
]

.pull-right[
<img src="animated-and-interactive-visualizations_files/figure-html/unnamed-chunk-4-1.gif" width="100%" style="display: block; margin: auto;" />
]

---

### Labeling

Labeling uses special **glue** syntax inside `labs()`.

```r
anim2 <- anim +
* labs(title = "Year: {frame_time}")
anim2
```

---

### Fine-tuning

Use `animate()` directly for more control over the animation.

```r
*animate(anim2, duration = 20, nframes = 200)
```

---

## Interactive Visualizations

Interactive visualization 
- works well with most datasets, especially complex ones, but is primarily suited for webpages.

- is often useful in tandem with carefully designed static visualizations.

.pull-left[
### plotly

Simple interactive visualizations can be achieved easily using [plotly](https://plotly.com/).

Two alternatives:

- call `plot_ly()` directly
- use `ggplotly()` on a plot built in ggplot2

We focus on the second alternative.
]

.pull-right[
<img src="images/plotly.png" width="204" style="display: block; margin: auto;" />
]

---

```r
library(plotly)
p <- ggplot(msleep, aes(brainwt, sleep_total, color = vore, 
                        genus = genus, 
                        conservation = conservation)) +
  geom_point() +
  scale_x_log10()
*ggplotly(p)
```

---

### Shiny

For more advanced interactive visualizations (beyond the scope of this course), 
have a look at [shiny](https://shiny.rstudio.com/).