Take A Sad Plot & Make It Better

class: center, middle, inverse, title-slide

# Take A Sad Plot & Make It Better
## A Case Study with R and ggplot2
### Alison Hill, PhD Data Scientist & Professional Educator <a href="https://github.com/apreshill">  apreshill</a> <a href="https://twitter.com/apreshill">   @apreshill</a> <a href="mailto:alison@rstudio.com">  alison@rstudio.com</a>

---

class: center, middle, inverse
## About me

![](images/beatles-tidy.gif)

---
class: center, middle

![](images/aim-here.png)

---
class: inverse, middle, center

### The prevalence puzzle: autism counts.

> Weintraub, K. (2011). The prevalence puzzle: autism counts. Nature News, 479(7371), 22-24.

.footnote[
Source: https://www.nature.com/news/2011/111102/full/479022a.html
]

---
class: inverse, top, center

### California Autism Prevalence Trends from 1931 to 2014

> Nevison, C., Blaxill, M. & Zahorodny, W. J Autism Dev Disord (2018) 48: 4103. https://doi.org/10.1007/s10803-018-3670-2

.footnote[
https://link.springer.com/article/10.1007/s10803-018-3670-2
]

---
class: center, bottom, inverse

# `v1.ppt`

## Hey, Jude, don't make it bad...

---
background-image: url("./images/version1.png")
background-size: cover

---

# Blameless postmortem

.pull-left[
![](images/version1.png)
]

.pull-right[
- 😀 Pink blobs are always smaller than green squares- always under-estimating.

- 😒 But, on R, there are more dots in the pink blob. The point is that more people are getting services, but...

- 😬 The "true" prevalence (total number of dots) looks different. It is actually the same at both time points!

- 😳 I actually like the dots! And I return to this later.
]

---
class: center, inverse, middle

# 💡

## What is the message?

Is there an autism epidemic? Based on the data available, we don't know!

---
class: center, inverse, middle

# 💡

## What is the medium?

A book chapter, but also presentations like this one!

---
class: center, middle

![](images/aim-here-v1.png)

---
class: center, bottom, inverse

# `v2.doc`

## Take a sad (plot) and make it better...

---
class: center
background-image: url("./images/version2.png")
background-size: 800px 550px

---
class: center, middle, inverse

![](images/beatles-face-palm.gif)

---

# Full-of-blame postmortem

.pull-left[
![](images/version2.png)
]

.pull-right[

- 😬 The connection between the line and pie charts is totally unclear. They are also kind of redundant.

- 😬 There are pie charts. With some kind of reflective coating.
]

---
class: center, middle

![](images/aim-here-v2.png)

---
class: center, bottom, inverse

# `v3 <- ggplot()`

## Then you'll begin to make it better...

---
class: center, middle, inverse

![](images/beatles-help.gif)

---
class: center, middle

# Asking for help

.pull-left[
![](images/help-me-viz-logo.jpg)
]

.pull-right[
Help Me Viz

[https://policyviz.com/helpmeviz/](https://policyviz.com/helpmeviz/)

*"A place to facilitate discussion, debate, and collaboration from the data visualization community"*
]

---
class: center, middle, inverse

![](images/help-me-newcat.png)

---
class: center
background-image: url("./images/help-me-plea.png")
background-size: auto

---
class: center
background-image: url("./images/help-me-response.png")
background-size: auto

---
class: center, inverse, middle

# 💡

## I can use `ggplot2`!

---

```r
set.seed(1000)
asdpop_base <- tibble::tibble(
 time1 = sample(1:100, 100, replace = F), 
 time2 = time1) %>% 
 tidyr::gather(x, y, time1:time2, factor_key = TRUE) 
asdpop_base
```

```
# A tibble: 200 x 2
 x y
 <fct> <int>
 1 time1 33
 2 time1 76
 3 time1 12
 4 time1 68
 5 time1 50
 6 time1 7
 7 time1 70
 8 time1 55
 9 time1 20
10 time1 24
# … with 190 more rows
```

---

```r
asdpop <- asdpop_base %>% 
 mutate(services = as.factor(case_when(
 x == "time1" & y <= 30 ~ 1, 
 x == "time1" & y > 30 ~ 0, 
 x == "time2" & y <= 60 ~ 1, 
 TRUE ~ 0
 )))
asdpop
```

```
# A tibble: 200 x 3
 x y services
 <fct> <int> <fct> 
 1 time1 33 0 
 2 time1 76 0 
 3 time1 12 1 
 4 time1 68 0 
 5 time1 50 0 
 6 time1 7 1 
 7 time1 70 0 
 8 time1 55 0 
 9 time1 20 1 
10 time1 24 1 
# … with 190 more rows
```

---
class: center

---
class: center

---
class: center

---
class: center

---
class: center

---
class: center
<img src="index_files/figure-html/bar6-1.png" width="80%" style="display: block; margin: auto;" />

---
class: center

---
class: center

---
class: center, middle

![](images/aim-here-v3.png)

---
class: center, bottom, inverse

# `rep("better", 40)`

## better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better, better...

---
class: center, middle, inverse

## Why I ❤️ `ggplot2`

![](images/beatles-rotation.gif)

---
class: center

---
class: inverse, middle, center

### Wes Anderson: Fantastic Mr. Fox!

```r
wes_palette("FantasticFox1")
```

.footnote[
https://github.com/karthik/wesanderson
]
---

---
class: inverse, middle, center

# Viridis: Magma

.footnote[
https://ggplot2.tidyverse.org/reference/scale_viridis.html
]

---

---
class: center

---
class: center

---
class: center

---
exclude: TRUE

---
class: center

---
class: center

---

---
class: center, middle, inverse

## Share your sad plot!

.pull-left[
![](images/beatles-hold-your-hand.gif)
]

.pull-right[
<blockquote class="twitter-tweet" data-lang="en">Last year I did a talk called &quot;Take a Sad Plot &amp; Make It Better&quot; 😭 Now I&#39;m looking for brave souls (<a href="https://twitter.com/hashtag/rstats?src=hash&amp;ref_src=twsrc%5Etfw">#rstats</a>, <a href="https://twitter.com/hashtag/python?src=hash&amp;ref_src=twsrc%5Etfw">#python</a>, <a href="https://twitter.com/hashtag/Excel?src=hash&amp;ref_src=twsrc%5Etfw">#Excel</a>!) to share their sad plots with me for a Sad Plot Showcase. No shame, just sharing! 🤝<a href="https://t.co/E0HNX1CRfU">https://t.co/E0HNX1CRfU</a><a href="https://twitter.com/hashtag/dataviz?src=hash&amp;ref_src=twsrc%5Etfw">#dataviz</a> <a href="https://twitter.com/hashtag/ggplot2?src=hash&amp;ref_src=twsrc%5Etfw">#ggplot2</a> <a href="https://twitter.com/hashtag/TidyTuesday?src=hash&amp;ref_src=twsrc%5Etfw">#TidyTuesday</a> <a href="https://t.co/T5ildPrDq0">pic.twitter.com/T5ildPrDq0</a>&mdash; Alison Hill (@apreshill) <a href="https://twitter.com/apreshill/status/1095535103139168256?ref_src=twsrc%5Etfw">February 13, 2019</a></blockquote>
<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>

]

---
class: center, middle, inverse

## Thanks!

Slides created via the R package [**xaringan**](https://github.com/yihui/xaringan).

The chakra comes from [remark.js](https://remarkjs.com), [**knitr**](http://yihui.name/knitr), and [R Markdown](https://rmarkdown.rstudio.com).

This work is licensed under a Creative Commons Attribution- ShareAlike 4.0 International License.

To view a copy of this license, visit:

https://creativecommons.org/licenses/by-sa/4.0/

---

Let's try gganimate

---

---