Slides 04 - Metaphors with Graphics

.title[
# Slides 04 - Metaphors with Graphics
]
.subtitle[
## From Code to Geometry
]
.author[
### Arvind Venkatadri
]
.institute[
### Srishti Manipal Institute
]
.date[
### (2022-08-09)
]

---

![shakespeare](https://media.giphy.com/media/oveqQA2LxpwYg/giphy.gif)

~~To code or not to code, that is the question...~~

---

# What is a Grammar of Graphics?

## Code looks and reads like **English**.   
## Has **verbs**, **nouns**, some **adjectives**....

- Describes Information/ideas/concepts from *any* **source domain**.

- **GEOMETRY** as the *target domain* : What comes out of R is predominantly "geometry"

---

--
.font120[
- **.hlb[Geom]etric objects** that appear on the plot
]

--
.font120[
- **.hlb[Aes]thetic mappings** from data to visual component
]

--
.font120[
- **.hlb[Stat]istics** transform data on the way to visualization
]

--
.font120[
- **.hlb[Coord]inates** organize location of geometric objects
]

--
.font120[
- **.hlb[Scale]s** define the range of values for aesthetics
]

--
.font120[
- **.hlb[Facet]s** group into subplots
]

---

# The Essence of ggplot

all `ggplot2`

- `aes(x = , y = )` (aesthetics)
- `aes(x = , y = , color = )` (add color)
- `aes(x = , y = , size = )` (add size)
- `+ facet_wrap(~ )` (facetting)
- `+ scale_` ( add a scale)

---
# gg is for Grammar of Graphics

```r
+ geom_*()
```
]

<img src="04-Metaphors-with-Graphics_files/figure-html/geom_demo-1.png" width="850px" height="350px" />
]

---

# [The Five-Named Graphs](http://moderndive.com/3-viz.html#FiveNG)

- Scatterplot: `geom_point()`
- Line graph: `geom_line()`
- Histogram: `geom_histogram()`
- Boxplot: `geom_boxplot()`
- Bar graph: `geom_bar()` or `geom_col` (see [Lab 02](../02-Pronouns-and-Data.html))

---

## Chunk : penguins

```r
head(penguins)
```

```
## # A tibble: 6 × 8
##   species island    bill_length_mm bill_depth_mm flipper_length_mm body_mass_g sex     year
##   <fct>   <fct>              <dbl>         <dbl>             <int>       <int> <fct>  <int>
*## 1 Adelie  Torgersen           39.1          18.7               181        3750 male    2007
*## 2 Adelie  Torgersen           39.5          17.4               186        3800 female  2007
*## 3 Adelie  Torgersen           40.3          18                 195        3250 female  2007
## 4 Adelie  Torgersen           NA            NA                  NA          NA <NA>    2007
## 5 Adelie  Torgersen           36.7          19.3               193        3450 female  2007
## 6 Adelie  Torgersen           39.3          20.6               190        3650 male    2007
```
We see the first few rows of the dataset `penguins`. We see that there are a few **NA** data observations too. Let us remove them for now.

```r
penguins <- penguins %>% drop_na()
```

---

## Chunk: Mapping

```r
*ggplot(penguins)
```
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/first-plot1a-out-1.png" width="504" />
]

---

## Chunk: Mapping

```r
ggplot(data = penguins, 
*      mapping = aes(x = bill_length_mm,
*          y = body_mass_g))
```
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/first-plot1b-out-1.png" width="504" />
]

---

## Chunk: Mapping

```r
ggplot(data = penguins, 
       mapping = aes(x = bill_length_mm, 
           y = body_mass_g)) +
* geom_point()
```
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/first-plot1c-out-1.png" width="504" />
]

---
## Chunk: Mapping

```r
ggplot(data = penguins, 
       mapping = aes(x = bill_length_mm, 
           y = body_mass_g)) +
  geom_point() +
* geom_smooth(method = "lm")
```
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/first-plot1d-out-1.png" width="504" />
]

---

### Chunk: Geom_Point_Position_Colour

```r
*ggplot(data = penguins)
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-2-1.png" width="504" />
]

---

### Chunk: Geom_Point_Position_Colour

```r
ggplot(data = penguins,
*      aes(x = bill_length_mm,
*          y = body_mass_g,
*          color = island))
```
We can leave out the "mapping" word and just use **aes** .

Why is there no plot?

🤔 💭

Right !! We have not used a `geom` command yet!!
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-3-1.png" width="504" />
]

---

### Chunk: Geom_Point_Position_Colour

```r
ggplot(data = penguins,
  aes(x = bill_length_mm,
      y = body_mass_g,
      color = island)) +  
*geom_point() +
* ggtitle("A point geom with position, color aesthetics")
```
Note that the points are located by **position** coordinates on both x and y axis, and **coloured** by the island variable.

]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-4-1.png" width="504" />
]

---

### Chunk: Geom_Point_Position_Colour

```r
ggplot(data = penguins,
  aes(x = bill_length_mm,
      y = body_mass_g,
      color = island)) +  
*geom_point(size = 4) +
* ggtitle("A point geom with position color and size aesthetics")
```
Note that the points are located by **position** coordinates on both x and y axis, and **coloured** by the island variable.

And we've fixed size = 4!

]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-5-1.png" width="504" />
]

---

## Alpha

```r
diamonds %>% 
  # Sample some 20% of the data
  slice_sample(prop = 0.2) %>% 
  ggplot(.) +
* geom_point(aes(x = carat,
*                y = price))
```

Are the points all overlapping?
Can we see them better?
]

.pull-right[
  
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-6-1.png" width="504" />
]

---

## Alpha

```r
diamonds %>% 
  # Sample some 20% of the data
  slice_sample(prop = 0.2) %>% 
  ggplot(.) +
* geom_point(aes(x = carat, y = price),
*            # alpha outside the aes() !!!
*            alpha = 0.2) +
  labs(title = "Points plotted with Alpha")
```

Are the points all overlapping?
Can we see them better?

]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-7-1.png" width="90%" />
]

---

## Chunk: Box Plot
.pull-left[

```r
ggplot(diamonds) + 
* geom_boxplot(aes(x = cut, y = price)) +
  labs(title = "Box Plot")
```

]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-8-1.png" width="90%" />
]

---

## Chunk: Box Plot
.pull-left[

```r
ggplot(diamonds) + 
* geom_boxplot(aes(x = cut,
*                  y = price,
*                  fill = cut)) +
  labs(title = "Box Plot")
```

]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-9-1.png" width="90%" />
]
---

## Chunk: Geom_Bar_1

```r
*ggplot(data = penguins)
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-10-1.png" width="90%" />
]

---

## Chunk: Geom_Bar_1

```r
ggplot(data = penguins) +  
*      aes(x = species)
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-11-1.png" width="90%" />
]

---

## Chunk: Geom_Bar_1

```r
ggplot(data = penguins) +  
       aes(x = species) +
* geom_bar() +
* ggtitle("A bar geom with position and height aesthetics")
```

The bars are plotted with **positions** on the x-axis, defined by the `species` variable,  and **heights** mapped to the y-axis.

How did the graph "know" the heights of the bars?

`geom_bar` has an internal `count` statistic computation. 
Many `geom_s` have internal computation that are accessible to programmers.

]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-12-1.png" width="90%" />
]

---

## Geom_Bar_Position_Stack_and_Dodge

.pull-left[
When using more than a pair of variables with a bar chart, we have a few more **position** options:

```r
  ggplot(penguins,
*          aes(x = species,
*              fill = island))
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-13-1.png" width="90%" />
]

---

## Geom_Bar_Position_Stack_and_Dodge

.pull-left[
When using more than a pair of variables with a bar chart, we have a few more **position** options:

```r
ggplot(penguins,
           aes(x = species,
               fill = island)) + 
* geom_bar() +
* ggtitle(label = "A stacked bar chart")
```

The bars are coloured by the `island` variable and are **stacked** in **position**.
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-14-1.png" width="90%" />
]
---

## Geom_Bar_Position_Stack_and_Dodge

```r
ggplot(penguins,
           aes(x = species,
               fill = island)) + 
* geom_bar(position ="dodge") +
* ggtitle(label =
*           "A dodged bar chart")
```
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-15-1.png" width="90%" />
]

---

## Facetting

```r
*ggplot(penguins)
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-16-1.png" width="90%" />
]

---

## Facetting

```r
ggplot(penguins) +
* aes(x = flipper_length_mm,
*     y = body_mass_g)
```
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-17-1.png" width="90%" />
]

---

## Facetting

```r
ggplot(penguins) +
  aes(x = flipper_length_mm,
      y = body_mass_g) +
* geom_point()
```
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-18-1.png" width="90%" />
]

---

## Facetting

```r
ggplot(penguins) +
  aes(x = flipper_length_mm,
      y = body_mass_g) +
  geom_point() +
* facet_wrap(~island) +
* ggtitle("A point geom graph with facets")
```

The graph has **split** into multiples, based on the **number** of islands. 
]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-19-1.png" width="90%" />
]
---

## Still more Facetting

```r
ggplot(penguins) +
  aes(x = flipper_length_mm,
      y = body_mass_g) +
* geom_point()
```

What if we have even more "factor" variables?
We have `island` and `species`...can we split further?

]

.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-20-1.png" width="90%" />
]

---

## Still more Facetting

```r
ggplot(penguins) +
  aes(x = flipper_length_mm,
      y = body_mass_g) +
  geom_point() +
* facet_grid(species~island) +
* ggtitle("A point geom graph with grid facets")
```

The graph has **split** into multiples, based on the **number** of islands **and** the number of species.

]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-21-1.png" width="90%" />
]

---

class:middle, center, inverse
## And shall we look briefly at colour?

---

## Finally...Colour !! ( Just a bit )

```r
diamonds %>% 
  slice_sample(prop = 0.2) %>% 
  ggplot(.) +
* geom_point(aes(x = carat, y = price))
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-22-1.png" width="90%" />
]
---

## Finally...Colour !! ( Just a bit )

```r
diamonds %>% 
  slice_sample(prop = 0.2) %>% 
  ggplot(.) +
  geom_point(aes(x = carat, y = price, colour = cut), size = 3) +
* scale_colour_brewer(palette = "Set3") +
  labs(title = "Brewer Colour Pallette (Set3)")
```
We are using the `RColorBrewer` package here. 
Type `RColorBrewer::display.brewer.all()` in your Console and see what palettes are available.

]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-23-1.png" width="90%" />
]

---

## Chunk: Colour !! ( Just a bit )

```r
diamonds %>% 
  slice_sample(prop = 0.2) %>% 
  ggplot(.) +
  geom_point(aes(x = carat, y = price, colour = cut), size = 3) +
*              scale_colour_viridis_d() +
  labs(title = "Viridis Palette",
       subtitle = "The Default in ggplot")
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-24-1.png" width="504" />
]

---

## Chunk: Colour !! ( Just a bit )

```r
diamonds %>% 
  slice_sample(prop = 0.2) %>% 
  ggplot(.) +
  geom_point(aes(x = carat, y = price, colour = cut), size = 3) +
*              scale_colour_viridis_d(option = "magma") +
* labs(title = "Viridis Palette, Option Magma")
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-25-1.png" width="504" />
]

---

## Chunk: Colour !! ( Just a bit )

```r
diamonds %>% 
  slice_sample(prop = 0.2) %>% 
  ggplot(.) +
  geom_point(aes(x = carat, y = price, colour = cut), size = 3) +
*              scale_colour_viridis_d(option = "inferno") +
  labs(title = "Viridis Palette, Option Inferno")
```
]
.pull-right[
<img src="04-Metaphors-with-Graphics_files/figure-html/unnamed-chunk-26-1.png" width="504" />
]

---

## Conclusion
- `ggplot` takes a dataframe/tibble as the data argument
- The `aes`-thetic arguments can be `x`, `y`, `colour`, `shape`, `alpha` for example...
- The `geom_*()` commands specify the kind of plot
- Together, the `ggplot` package offers a **Grammar** of near-English commands which allow us to plot data in various ways.

---

# References

1. [Wickham, Hadley. (2010) "A Layered Grammar of Graphics". *Journal of Computational and Graphical Statistics*, 19(1).](http://www.jstor.org.proxy.uchicago.edu/stable/25651297)

2. [Wilkinson, Leland. (2005). *The Grammar of Graphics*. (UChicago authentication required)](http://link.springer.com.proxy.uchicago.edu/book/10.1007%2F0-387-28695-0)