Optimizing color spaces

---

# Agenda / heads up:

```r
# install if necessary
install.packages(c("tidyverse", "here", "colorspace",
                   "scales", "ggthemes", "usethis", "cowplot"))

usethis::use_course("MACS40700/choosing-colors")
```

FYI for some plots, you may need to find them on their own and install separately, e.g.: https://github.com/UrbanInstitute/urbnmapr

---

# Uses of color in data visualization

<table style = "border: none; line-height: 2.5;">
<tr style = "background: white;">
<td style = "text-align: left; width: 50%;">
1. Distinguish categories (qualitative)
</td>
<td>
<img src = "images/qualitative.png" width = 100% style = "text-align: right; vertical-align: middle"></img>
</td>
</tr>
</table>

---

# Qualitative scale example

Palette name: Okabe-Ito

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Qualitative scale example

Palette name: ColorBrewer Set1

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Qualitative scale example

Palette name: ColorBrewer Set3

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Uses of color in data visualization

<tr style = "background: white;">
<td style = "text-align: left;">
2. Represent numeric values (sequential)
</td>
<td>
<img src = "images/sequential.png" width = 100% style = "text-align: right; vertical-align: middle"></img>
</td>
</tr>
</table>

---

# Sequential scale example

Palette name: Viridis

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Sequential scale example

Palette name: Inferno

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Sequential scale example

Palette name: Cividis

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Uses of color in data visualization

<tr style = "background: white;">
<td style = "text-align: left;">
3. Represent numeric values (diverging)
</td>
<td>
<img src = "images/diverging.png" width = 100% style = "text-align: right; vertical-align: middle"></img>
</td>
</tr>
</table>

---

# Diverging scale example

Palette name: ColorBrewer PiYG

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Diverging scale example

Palette name: Carto Earth

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Diverging scale example

Palette name: Blue-Red

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

# Uses of color in data visualization

<tr style = "background: white;">
<td style = "text-align: left;">
4. Highlight
</td>
<td>
<img src = "images/highlight.png" width = 100% style = "text-align: right; vertical-align: middle"></img>
</td>
</tr>
</table>

---

# Highlight example

.panelset[
.panel[.panel-name[Gray]
<img src="index_files/figure-html/Aus-athletes-track-1.png" width="70%" style="display: block; margin: auto;" />

Palette name: Grays with accents

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

]
.panel[.panel-name[Gray2]
<img src="index_files/figure-html/grays-1.png" width="70%" style="display: block; margin: auto;" />

Palette name: Grays with accents

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

]

.panel[.panel-name[OkabeIto]
<img src="index_files/figure-html/Aus-athletes-track2-1.png" width="70%" style="display: block; margin: auto;" />

Palette name: Okabe-Ito accent

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)
]
.panel[.panel-name[Accent]

Palette name: ColorBrewer accent

???

Figure redrawn from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)
]]
---

# Uses of color in data visualization

---

# Choosing a color scale

---

# Choosing a color scale

- Emphasis on interpretability and accessibility
- Default palettes are less than desirable
- Variables may require transformations

---

# Default palette in `ggplot2`

---

# Suboptimal default choices

---

# Common forms of color vision deficiency

### Red-green

- Deuteranomaly
- Protanomaly
- Protanopia and deuteranopia

### Blue-yellow

- Tritanomaly
- Tritanopia

### Complete color vision deficiency

- Monochromacy

---

# Inspecting for color vision deficiency

---

# Inspecting for color vision deficiency

``` r
library(colorblindr) # https://www.rdocumentation.org/packages/colorblindr/versions/0.1.0 
cvd_grid(plot = pen_fig)
```

---

# Inspecting for color deficiency

---

# Inspecting for color deficiency

---

# When to use quantitative or qualitative color scales?

---

# Use qualitative for nominal variables

---

# Use quantitative for numerical variables

---

# Quantitative `\(\neq\)` continuous

---

# Shades to emphasize order

---

# Double-encoded line chart

---

# Implementing optimal color palettes in R

---

## **ggplot2** color scale functions

.small.center[

---

## **ggplot2** color scale functions are a bit of a mess

.small.center[

---

## **ggplot2** color scale functions are a bit of a mess

.small.center[

---

## **ggplot2** color scale functions are a bit of a mess

.small.center[

---

## **ggplot2** color scale functions are a bit of a mess

.small.center[

... and there are many many more

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles1-out-1.png" width="80%" style="display: block; margin: auto;" />
]

``` r
ggplot(temps_months, aes(x = month, y = location, fill = mean)) + 
  geom_tile(width = 0.95, height = 0.95) + 
  coord_fixed(expand = FALSE) +
  theme_classic()
  # no fill scale defined, default is scale_fill_gradient()
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles2-out-1.png" width="80%" style="display: block; margin: auto;" />
]

``` r
ggplot(temps_months, aes(x = month, y = location, fill = mean)) + 
  geom_tile(width = 0.95, height = 0.95) + 
  coord_fixed(expand = FALSE) +
  theme_classic() +
* scale_fill_gradient()
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles3-out-1.png" width="80%" style="display: block; margin: auto;" />
]

``` r
ggplot(temps_months, aes(x = month, y = location, fill = mean)) + 
  geom_tile(width = 0.95, height = 0.95) + 
  coord_fixed(expand = FALSE) +
  theme_classic() +
* scale_fill_viridis_c()
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles4-out-1.png" width="80%" style="display: block; margin: auto;" />
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles5-out-1.png" width="80%" style="display: block; margin: auto;" />
]

``` r
ggplot(temps_months, aes(x = month, y = location, fill = mean)) + 
  geom_tile(width = 0.95, height = 0.95) + 
  coord_fixed(expand = FALSE) +
  theme_classic() +
* scale_fill_distiller(palette = "YlGnBu")
```
]
]

---

## The `colorspace` package creates some order

Scale name: `scale_<aesthetic>_<datatype>_<colorscale>()`

- `<aesthetic>`: name of the aesthetic (`fill`, `color`, `colour`)
- `<datatype>`: type of variable plotted (`discrete`, `continuous`, `binned`)
- `<colorscale>`: type of the color scale (`qualitative`, `sequential`, `diverging`, `divergingx`)

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles6-out-1.png" width="80%" style="display: block; margin: auto;" />
]

``` r
ggplot(temps_months, aes(x = month, y = location, fill = mean)) + 
  geom_tile(width = 0.95, height = 0.95) + 
  coord_fixed(expand = FALSE) +
  theme_classic() +
* scale_fill_continuous_sequential(palette = "YlGnBu")
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles7-out-1.png" width="80%" style="display: block; margin: auto;" />
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/temps-tiles8-out-1.png" width="80%" style="display: block; margin: auto;" />
]

---

``` r
colorspace::hcl_palettes(type = "sequential", plot = TRUE) # all sequential palettes
```

---

``` r
colorspace::hcl_palettes(type = "diverging", plot = TRUE, n = 9) # all diverging palettes
```

---

``` r
colorspace::divergingx_palettes(plot = TRUE, n = 9) # all divergingx palettes
```

---

# Setting colors for discrete, qualitative scales

---

``` r
colorspace::hcl_palettes(type = "qualitative", plot = TRUE) # all qualitative palettes
```

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/qual-scales-example1-out-1.png" width="70%" style="display: block; margin: auto;" />
]

``` r
ggplot(popgrowth, aes(x = pop2000, y = popgrowth, color = region)) +
  geom_point(size = 4) +
  scale_x_log10()
  # no color scale defined, default is scale_color_hue()
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/qual-scales-example2-out-1.png" width="70%" style="display: block; margin: auto;" />
]

``` r
ggplot(popgrowth, aes(x = pop2000, y = popgrowth, color = region)) +
  geom_point(size = 4) +
  scale_x_log10() +
* scale_color_hue()
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/qual-scales-example3a-out-1.png" width="70%" style="display: block; margin: auto;" />
]

``` r
ggplot(popgrowth, aes(x = pop2000, y = popgrowth, color = region)) +
  geom_point(size = 4) +
  scale_x_log10() +
  # uses Pastel 1
* scale_color_discrete_qualitative(palette = "Dark 2")
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/qual-scales-example3-out-1.png" width="70%" style="display: block; margin: auto;" />
]

``` r
library(ggthemes)  # for scale_color_colorblind()

ggplot(popgrowth, aes(x = pop2000, y = popgrowth, color = region)) +
  geom_point(size = 4) +
  scale_x_log10() +
  # uses Okabe-Ito colors
* scale_color_colorblind()
```
]
]

---

## Examples

.panelset[
.panel[.panel-name[Output]
<img src="index_files/figure-html/qual-scales-example4-out-1.png" width="70%" style="display: block; margin: auto;" />
]

``` r
ggplot(popgrowth, aes(x = pop2000, y = popgrowth, color = region)) +
  geom_point(size = 4) +
  scale_x_log10() +
* scale_color_manual(
*   values = c(
*     West = "#E69F00", South = "#56B4E9",
*     Midwest = "#009E73", Northeast = "#F0E442"
*   )
* )
```
]
]

---

## Okabe-Ito RGB codes

.center[
<img src = "https://clauswilke.com/dataviz/pitfalls_of_color_use_files/figure-html/palette-Okabe-Ito-1.png", width = 100%></img>
]

.tiny-font[
Name           | Hex code &nbsp;&nbsp; | R, G, B (0-255)
:----------    | :-------  | :--------
orange         | #E69F00   | 230, 159, 0
sky blue	     | #56B4E9   | 86, 180, 233
bluish green   | #009E73   | 0, 158, 115
yellow	       | #F0E442   | 240, 228, 66
blue	         | #0072B2   | 0, 114, 178
vermilion	     | #D55E00   | 213, 94, 0
reddish purple | #CC79A7   | 204, 121, 167
black	         | #000000   | 0, 0, 0
]

???

Figure from [Claus O. Wilke. Fundamentals of Data Visualization. O'Reilly, 2019.](https://clauswilke.com/dataviz)

---

## Apply it -- work in groups through the exercises here

```r
# install if necessary
install.packages(c("tidyverse", "here", "colorspace",
                   "scales", "ggthemes", "usethis", "cowgrid"))

usethis::use_course("MACS40700/choosing-colors")
```

---
# Recap

* Consider the palettes: balance saturation
* Colors: 
  * Qualitative: diverging
  * Quantitative: scale