Modifiers

# Modifiers
### Gina Reynolds, July 2019

---

---

# Modifiers

> *"The book is on the table"*

> *"The huge, paperback book is on the purple table"?*

---
class: center, middle

We are able to learn something about our data with very few inputs with ggplot2:

declare data -- `ggplot(data = ?)`

declare aesthetic representation -- `aes(x = ?)`

declare geom  -- `geom_rug()`

... and we get a plot back.

---
class: center, middle

Part of the reason we have such a concise "grammar of graphics sentence" is there are convenient and reasonable defaults related to the aesthetics.

In the first human, "The book is on the table", we also might a picture of the book and table -- it is some reasonable "default image" of books -- a genaric book, a genaric table.

But we might use modifiers to change the picture, to specify that the table is not *brown* as we might imagine by default, but *purple*!

---
class: center, middle

The same is true in ggplot2, also, we can write short "sentences" and get a meaningful plot in return because of helpful defaults.  But we can use modifiers to change these defaults.

---

# Modifiers

Modifiers adjust the default associated with aesthetic representations for:

1. Easy: the scale *labels* (the scale labels are the variable names by default, but they could be something else).  *Mastry is fairly easy. *

2. Medium: the *coordinate system* (the default coordinate system is cartesian (x horizontal, y vertical), but it could be something else - like polar).  *There's a bit more to know here.*

3. Hard: the aesthetic scales *themselves* (we have seen a "blues" scale by default for the continuous color aesthetic, but it could be greens, or yellows). *There's a lot to learn here!*

---

## Modifiers I: aesthetic scale labels

Aesthetic scales also get default *labels* which are the variable names.  We can also modify these.

The syntax is labs(\* = "New label"), where the \* is the aesthetic name.

`labs()` can also be used to add plot labels, plot title, caption, subtitle and tag.

Let's see how.

---

```r
*ggplot(data = gapminder_2002) +
* aes(x = gdpPercap) +
* aes(y = lifeExp) +
* geom_point() +
* aes(color = continent) +
* aes(shape = continent) +
* aes(size = gdpPercap) +
* aes(alpha = gdpPercap)   # this much is familiar
```
]]
.column[.content[
<img src="modifiers_files/figure-html/labs_user_1_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  aes(color = continent) +  
  aes(shape = continent) +  
  aes(size = gdpPercap) +  
  aes(alpha = gdpPercap) +  # this much is familiar 
* labs(x = "GDP per cap")
```
]]
.column[.content[
<img src="modifiers_files/figure-html/labs_user_2_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  aes(color = continent) +  
  aes(shape = continent) +  
  aes(size = gdpPercap) +  
  aes(alpha = gdpPercap) +  # this much is familiar 
  labs(x = "GDP per cap") +  
  labs(y = "Life Expectency") +  
  labs(color = "Continent") +  
  labs(shape = "Continent") +  
  labs(size = "GDP per cap") +  
  labs(alpha = "GDP per cap") +  
* # Note: labs is also used to adjust *plot* labels  # Note: labs is also used to adjust *plot* labels
* labs(title = "GDP per Cap versus\nLife Expectancy in 2002")
```
]]
.column[.content[
<img src="modifiers_files/figure-html/labs_user_8_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  aes(color = continent) +  
  aes(shape = continent) +  
  aes(size = gdpPercap) +  
  aes(alpha = gdpPercap) +  # this much is familiar 
  labs(x = "GDP per cap") +  
  labs(y = "Life Expectency") +  
  labs(color = "Continent") +  
  labs(shape = "Continent") +  
  labs(size = "GDP per cap") +  
  labs(alpha = "GDP per cap") +  
  # Note: labs is also used to adjust *plot* labels  # Note: labs is also used to adjust *plot* labels
  labs(title = "GDP per Cap versus\nLife Expectancy in 2002") +  
* labs(subtitle = "Data: Gapminder Data in R")
```
]]
.column[.content[
<img src="modifiers_files/figure-html/labs_user_9_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  aes(color = continent) +  
  aes(shape = continent) +  
  aes(size = gdpPercap) +  
  aes(alpha = gdpPercap) +  # this much is familiar 
  labs(x = "GDP per cap") +  
  labs(y = "Life Expectency") +  
  labs(color = "Continent") +  
  labs(shape = "Continent") +  
  labs(size = "GDP per cap") +  
  labs(alpha = "GDP per cap") +  
  # Note: labs is also used to adjust *plot* labels  # Note: labs is also used to adjust *plot* labels
  labs(title = "GDP per Cap versus\nLife Expectancy in 2002") +  
  labs(subtitle = "Data: Gapminder Data in R") +  
* labs(caption = "Visualization: @EvaMaeRey")
```
]]
.column[.content[
<img src="modifiers_files/figure-html/labs_user_10_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  aes(color = continent) +  
  aes(shape = continent) +  
  aes(size = gdpPercap) +  
  aes(alpha = gdpPercap) +  # this much is familiar 
  labs(x = "GDP per cap") +  
  labs(y = "Life Expectency") +  
  labs(color = "Continent") +  
  labs(shape = "Continent") +  
  labs(size = "GDP per cap") +  
  labs(alpha = "GDP per cap") +  
  # Note: labs is also used to adjust *plot* labels  # Note: labs is also used to adjust *plot* labels
  labs(title = "GDP per Cap versus\nLife Expectancy in 2002") +  
  labs(subtitle = "Data: Gapminder Data in R") +  
  labs(caption = "Visualization: @EvaMaeRey") +  
* labs(tag = "Plot 1")
```
]]
.column[.content[
<img src="modifiers_files/figure-html/labs_user_11_output-1.png" width="100%" />
]]

---

## Questions:

- When might you use labs(tag = )?
- What does the character string "\n" do in the place of a space in the plot title?
- Color and shape variables represent the same variable, and labs(color = \*) labs(shape = \*) have the same input, "Continent".  What does this mean for the legend? Change the labels to be different and see.
- Change the y label to an empty string, '""'.  What does this do?

---

## Modifiers II: x and y position coordinate system

A third default for how aesthetic mapping of *positions x and y* are manifest is the coordinate system.  The default is cartesian coordinates - where x lies on the horizontal and is pependicular to y.

But let's quickly check out some of the different options for coordinates systems.  When we use a coord_* statement in ggplot, we can overwrite the default.

---

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  coord_cartesian()   
```
]]
.column[.content[
<img src="modifiers_files/figure-html/coords_rotate_1_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
* coord_equal()
```
]]
.column[.content[
<img src="modifiers_files/figure-html/coords_rotate_2_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
* coord_fixed(ratio = 400)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/coords_rotate_3_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
* coord_flip()
```
]]
.column[.content[
<img src="modifiers_files/figure-html/coords_rotate_4_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
* coord_polar()
```
]]
.column[.content[
<img src="modifiers_files/figure-html/coords_rotate_5_output-1.png" width="100%" />
]]

---

## Modifiers III: aesthetic mapping (representation) scales

With the aes() statement we have declared a variable mapped to an aesthetic, like x position, color, line width, etc.  To go back and ask for a modified scale, we use scales\_\* statements.  These statements provide a quick way to update our preferences for all the scales used for mapped aesthetics.

Let's look at how we modify the default scale with a scales\_\* statement.

---

### Scales statements

There are a variety of scaling alternatives for the x position aesthetic:

- scale_x_log10()
- scale_x_reverse()
- scale_x_sqrt()

---

```r
*ggplot(data = gapminder_2002) +
* aes(x = gdpPercap) +
* aes(y = lifeExp) +
* geom_point()   # this much is familiar
```
]]
.column[.content[
<img src="modifiers_files/figure-html/x_scales_user_1_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  # this much is familiar 
* scale_x_log10()   # overrides default
```
]]
.column[.content[
<img src="modifiers_files/figure-html/x_scales_user_2_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  # this much is familiar 
  scale_x_log10() +  # overrides default  
* scale_x_reverse()   # overrides log10
```
]]
.column[.content[
<img src="modifiers_files/figure-html/x_scales_user_3_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  # this much is familiar 
  scale_x_log10() +  # overrides default  
  scale_x_reverse() +  # overrides log10  
* scale_x_sqrt()  # overrides reverse
```
]]
.column[.content[
<img src="modifiers_files/figure-html/x_scales_user_4_output-1.png" width="100%" />
]]

---

### More aesthetic scale modifications

*Note that using multiple scales associated with a single aesthetic is not typical.  We are just doing it here to expose you quickly to a variety of possibilities.  ggplot2 uses whatever scale specification last declared in our examples.*

Analogously, we can use scale\_\*\_\* statements to modify other aesthetic scales.  Here are some examples

- scale_x_log10() 
- scale_y_reverse(limits = c(100, 0)) 
- scale_color_viridis_d() 
- scale_shape_discrete(solid = F) 
- scale_size_continuous(breaks = c(100, 1000, 10000, 100000)) 
- scale_alpha(range = c(.5, 1)) # full transparency is 0

Let's tour several of those to get a taste.

---

```r
*ggplot(data = gapminder_2002)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_1_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
* aes(x = gdpPercap)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_2_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
* scale_x_continuous(trans = "log10")
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_3_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
* aes(y = lifeExp)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_4_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
* scale_y_continuous(breaks = c(50,70))
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_5_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
* geom_point()
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_6_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
* aes(color = continent)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_7_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
  aes(color = continent) +  
* scale_color_viridis_d()
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_8_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
  aes(color = continent) +  
  scale_color_viridis_d() +  
* aes(size = gdpPercap)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_9_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
  aes(color = continent) +  
  scale_color_viridis_d() +  
  aes(size = gdpPercap) +  
* scale_size_continuous(limits = c(0, 400000))
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_10_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
  aes(color = continent) +  
  scale_color_viridis_d() +  
  aes(size = gdpPercap) +  
  scale_size_continuous(limits = c(0, 400000)) +  
* aes(shape = continent)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_11_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
  aes(color = continent) +  
  scale_color_viridis_d() +  
  aes(size = gdpPercap) +  
  scale_size_continuous(limits = c(0, 400000)) +  
  aes(shape = continent) +  
* scale_shape_discrete(solid = F)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_12_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
  aes(color = continent) +  
  scale_color_viridis_d() +  
  aes(size = gdpPercap) +  
  scale_size_continuous(limits = c(0, 400000)) +  
  aes(shape = continent) +  
  scale_shape_discrete(solid = F) +  
* aes(alpha = gdpPercap)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_13_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  scale_x_continuous(trans = "log10") +  
  aes(y = lifeExp) +  
  scale_y_continuous(breaks = c(50,70)) +  
  geom_point() +  
  aes(color = continent) +  
  scale_color_viridis_d() +  
  aes(size = gdpPercap) +  
  scale_size_continuous(limits = c(0, 400000)) +  
  aes(shape = continent) +  
  scale_shape_discrete(solid = F) +  
  aes(alpha = gdpPercap) +  
* scale_alpha(range = c(.4, .9))
```
]]
.column[.content[
<img src="modifiers_files/figure-html/scales_auto_14_output-1.png" width="100%" />
]]

---

## Questions:

- What does the parameter "breaks" seem to control?

- What are some additional color scales that can be used?  Use auto complete in RStudio.  (i.e. type "scale_color" and then "tab" to see your options)

- What does "d" stand for in the scale_color_viridis_d()? Use the help pane.

- While you are in the help pane, see how would you set missing values (NA) to "black".

---

For a full tour of scale modifications check out the [scales second look](https://evamaerey.github.io/ggplot2_grammar_guide/scales).

---

## Modifiers - Review

We have seen that there are three modifiers to aesthetic mapping.

- scale_*
- labs(* = )
- coords_? (applies x,y position aesthetics only)

To review, let's just look at modifiers for the horizontal *x* position aesthetic.

---

```r
*ggplot(data = gapminder_2002)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/modifiers_review_auto_1_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
* aes(x = gdpPercap)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/modifiers_review_auto_2_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
* aes(y = lifeExp)
```
]]
.column[.content[
<img src="modifiers_files/figure-html/modifiers_review_auto_3_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
* geom_point()
```
]]
.column[.content[
<img src="modifiers_files/figure-html/modifiers_review_auto_4_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
* labs(x = "GDP per cap")   # modify x label
```
]]
.column[.content[
<img src="modifiers_files/figure-html/modifiers_review_auto_5_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  labs(x = "GDP per cap") +  # modify x label 
* coord_polar()   # modify x and y coordinate system
```
]]
.column[.content[
<img src="modifiers_files/figure-html/modifiers_review_auto_6_output-1.png" width="100%" />
]]
---
class: split-40
count: false

```r
ggplot(data = gapminder_2002) +  
  aes(x = gdpPercap) +  
  aes(y = lifeExp) +  
  geom_point() +  
  labs(x = "GDP per cap") +  # modify x label 
  coord_polar() +  # modify x and y coordinate system
* scale_x_log10()  # modify x scale
```
]]
.column[.content[
<img src="modifiers_files/figure-html/modifiers_review_auto_7_output-1.png" width="100%" />
]]