Deep dive: geometric objects (AKA geoms)

.title[
# Deep dive: geometric objects (AKA geoms)
]
.author[
### MACS 40700 <br /> University of Chicago
]

---

# Agenda:

* Aesthetic mappings: WHERE TO PUT!
* Geoms: making a layer cake
* Scales

---

## Announcements
* Assignment deadlines (often BUT NOT ALWAYS Weds / Fridays)

---

# Setup

---

## Packages

``` r
# load packages
library(tidyverse)
# remotes::install_github("MACS40700/c3s2datasets")
library(c3s2datasets)
```

---

## ggplot2 theme

``` r
# set default theme for ggplot2
ggplot2::theme_set(ggplot2::theme_minimal(base_size = 16))
```

---

## Figure sizing

For more on including figures in R Markdown documents with the right size, resolution, etc. the following resources are great:

- [R for Data Science - Graphics for communication](https://r4ds.had.co.nz/graphics-for-communication.html)
- [Tips and tricks for working with images and figures in R Markdown documents](https://www.zevross.com/blog/2017/06/19/tips-and-tricks-for-working-with-images-and-figures-in-r-markdown-documents/)

``` r
# set default figure parameters for knitr
knitr::opts_chunk$set(
  fig.width = 8, # 8"
  fig.asp = 0.618, # the golden ratio
  fig.retina = 2, # dpi multiplier for displaying HTML output on retina
  dpi = 150, # higher dpi, sharper image
  out.width = "70%"
)
```

---

## Data prep: new variables

From last class...

``` r
scorecard <- scorecard %>%
  mutate(pctpell_cat = cut_interval(x = pctpell, n = 6)) %>%
  drop_na(pctpell_cat)

scorecard %>%
  count(pctpell_cat)
```

```
## # A tibble: 6 × 2
##   pctpell_cat       n
##   <fct>         <int>
## 1 [0,0.163]       164
## 2 (0.163,0.326]   657
## 3 (0.326,0.489]   591
## 4 (0.489,0.652]   209
## 5 (0.652,0.816]    68
## 6 (0.816,0.979]    26
```

---

## Data prep: summary table

From last class...

``` r
mean_netcost_pctpell <- scorecard %>%
  group_by(pctpell_cat) %>%
  summarise(mean_netcost = mean(netcost, na.rm = TRUE))

mean_netcost_pctpell
```

```
## # A tibble: 6 × 2
##   pctpell_cat   mean_netcost
##   <fct>                <dbl>
## 1 [0,0.163]           28713.
## 2 (0.163,0.326]       21525.
## 3 (0.326,0.489]       18660.
## 4 (0.489,0.652]       17087.
## 5 (0.652,0.816]       14771.
## 6 (0.816,0.979]        8466.
```

---

# Aesthetic mappings in ggplot2

---

## Global vs. layer-specific aesthetics

- Aesthetic mappings can be supplied in the initial `ggplot()` call, in individual layers, or in some combination of both.

- Within each layer, you can add, override, or remove mappings.

- If you only have one layer in the plot, the way you specify aesthetics doesn’t make any difference. However, the distinction is important when you start adding additional layers.

---

## Activity: Spot the difference I
.task[
Do you expect the following plots to be the same or different? If different, how? Discuss in a pair (or group) without running the code and sketch the resulting plots based on what you think the code will produce.
]

``` r
# Plot A
ggplot(scorecard, aes(x = avgfacsal, y = netcost, color = pctpell_cat)) +
  geom_point(alpha = 0.7) +
  geom_smooth(method = "lm", se = FALSE, size = 0.5)
```

``` r
# Plot B
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(aes(color = pctpell_cat), alpha = 0.7) +
  geom_smooth(method = "lm", se = FALSE, size = 0.5)
```

<countdown-timer class="countdown" id="timer_f1c981c0" minutes="3" seconds="0" update-every="1" tabindex="0" style="right:0;bottom:0;"></countdown-timer>

---

## Activity: Spot the difference I

``` r
# Plot A
ggplot(scorecard, aes(x = avgfacsal, y = netcost, color = pctpell_cat)) +
  geom_point(alpha = 0.7) +
  geom_smooth(method = "lm", se = FALSE, size = 0.5)
```

<img src="index_files/figure-html/unnamed-chunk-8-1.png" width="60%" style="display: block; margin: auto;" />
] 
.panel[.panel-name["Plot B"]

``` r
# Plot B
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(aes(color = pctpell_cat), alpha = 0.7) +
  geom_smooth(method = "lm", se = FALSE, size = 0.5)
```

<img src="index_files/figure-html/unnamed-chunk-9-1.png" width="60%" style="display: block; margin: auto;" />
]]

---

## Activity: Spot the difference II

.task[
Do you expect the following plots to be the same or different? If different, how? Discuss in a pair (or group) without running the code and sketch the resulting plots based on what you think the code will produce.
]

``` r
# Plot A
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(aes(color = pctpell_cat))
```

``` r
# Plot B
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(color = "blue")
```

``` r
# Plot C
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(color = "#a493ba")
```

<countdown-timer class="countdown" id="timer_021a64ce" minutes="3" seconds="0" update-every="1" tabindex="0" style="right:0;bottom:0;"></countdown-timer>

---

## Activity: Spot the difference II

``` r
# Plot A
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(aes(color = pctpell_cat))
```

<img src="index_files/figure-html/unnamed-chunk-12-1.png" width="60%" style="display: block; margin: auto;" />
]

``` r
# Plot B
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(color = "blue")
```

<img src="index_files/figure-html/unnamed-chunk-13-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name["Plot C"]

``` r
# Plot C
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(color = "#800000")
```

<img src="index_files/figure-html/unnamed-chunk-14-1.png" width="60%" style="display: block; margin: auto;" />
]
]
---

# Geoms

---

## Geoms

- Geometric objects, or geoms for short, perform the actual rendering of the layer, controlling the type of plot that you create

- You can think of them as "the geometric shape used to represent the data"

---

## One variable

- Discrete:
  - `geom_bar()`: display distribution of discrete variable.

- Continuous
  - `geom_histogram()`: bin and count continuous variable, display with bars
  - `geom_density()`: smoothed density estimate
  - `geom_dotplot()`: stack individual points into a dot plot
  - `geom_freqpoly()`: bin and count continuous variable, display with lines

---

## .hand[aside...]

Always use "typewriter text" (monospace font) when writing function names, and follow with `()`, e.g.,

- `geom_freqpoly()`
- `mean()`
- `lm()`

---

## `geom_bar()`

``` r
ggplot(scorecard, aes(x = pctpell_cat)) +
  geom_bar()
```

---

## `geom_bar()`

``` r
ggplot(scorecard, aes(y = pctpell_cat)) +
  geom_bar()
```

---

## `geom_histogram()`

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_histogram()
```

---

## `geom_histogram()` and `binwidth`

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_histogram(binwidth = 2000)
```

<img src="index_files/figure-html/unnamed-chunk-18-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[5K]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_histogram(binwidth = 5000)
```

<img src="index_files/figure-html/unnamed-chunk-19-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[10K]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_histogram(binwidth = 10000)
```

<img src="index_files/figure-html/unnamed-chunk-20-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[20K]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_histogram(binwidth = 20000)
```

<img src="index_files/figure-html/unnamed-chunk-21-1.png" width="60%" style="display: block; margin: auto;" />
]
]

---

## `geom_density()`

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density()
```

---

## `geom_density()` and bandwidth (`bw`)

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(bw = 1)
```

<img src="index_files/figure-html/unnamed-chunk-23-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[100]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(bw = 100)
```

<img src="index_files/figure-html/unnamed-chunk-24-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[1000]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(bw = 1000)
```

<img src="index_files/figure-html/unnamed-chunk-25-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[10000]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(bw = 10000)
```

<img src="index_files/figure-html/unnamed-chunk-26-1.png" width="60%" style="display: block; margin: auto;" />
]
]

---

## `geom_density()` outlines

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(outline.type = "full")
```

<img src="index_files/figure-html/unnamed-chunk-27-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[both]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(outline.type = "both")
```

<img src="index_files/figure-html/unnamed-chunk-28-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[upper]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(outline.type = "upper")
```

<img src="index_files/figure-html/unnamed-chunk-29-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[lower]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_density(outline.type = "lower")
```

<img src="index_files/figure-html/unnamed-chunk-30-1.png" width="60%" style="display: block; margin: auto;" />
]
]

---

## `geom_dotplot()`

.task[
What does each point represent? How are their locations determined? What do the x and y axes represent?
]

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_dotplot(binwidth = 418)
```

<countdown-timer class="countdown" id="timer_e6e8f6ed" minutes="3" seconds="0" update-every="1" tabindex="0" style="right:0;bottom:0;"></countdown-timer>

---

## `geom_freqpoly()`

``` r
ggplot(scorecard, aes(x = netcost)) +
  geom_freqpoly(binwidth = 1000)
```

---

## `geom_freqpoly()` for comparisons

``` r
ggplot(scorecard, aes(x = netcost, fill = pctpell_cat)) +
  geom_histogram(binwidth = 5000)
```

<img src="index_files/figure-html/unnamed-chunk-34-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[Frequency polygon]

``` r
ggplot(scorecard, aes(x = netcost, color = pctpell_cat)) +
  geom_freqpoly(binwidth = 5000, size = 1)
```

<img src="index_files/figure-html/unnamed-chunk-35-1.png" width="60%" style="display: block; margin: auto;" />
]

``` r
ggplot(scorecard, aes(x = netcost, fill = pctpell_cat)) +
  geom_histogram(binwidth = 5000, position = "identity")
```

<img src="index_files/figure-html/unnamed-chunk-36-1.png" width="60%" style="display: block; margin: auto;" />
]

``` r
ggplot(scorecard, aes(x = netcost, fill = pctpell_cat)) +
  geom_histogram(binwidth = 5000, position = "identity", alpha = 0.5)
```

<img src="index_files/figure-html/unnamed-chunk-37-1.png" width="60%" style="display: block; margin: auto;" />
]
]

---

## Two variables - both continuous

- `geom_point()`: scatterplot
- `geom_quantile()`: smoothed quantile regression
- `geom_rug()`: marginal rug plots
- `geom_smooth()`: smoothed line of best fit
- `geom_text()`: text labels

---

## `geom_rug()`

``` r
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point() +
  geom_rug(alpha = 0.1)
```

---

## `geom_rug()` on the outside

``` r
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point() +
  geom_rug(alpha = 0.1, outside = TRUE) +
  coord_cartesian(clip = "off")
```

---

## `geom_rug()` on the outside, but better

``` r
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point() +
  geom_rug(alpha = 0.1, outside = TRUE, sides = "tr") +
  coord_cartesian(clip = "off") +
  theme(plot.margin = margin(1, 1, 1, 1, "cm"))
```

---

## `geom_text()`

``` r
sample_n(scorecard, size = 50) %>%
  ggplot(aes(x = avgfacsal, y = netcost)) +
  geom_text(aes(label = type))
```

---

## `geom_text()` and more

``` r
sample_n(scorecard, size = 50) %>%
  ggplot(aes(x = avgfacsal, y = netcost)) +
  geom_text(aes(label = type, color = type))
```

---

## `geom_text()` and even more

``` r
sample_n(scorecard, size = 50) %>%
  ggplot(aes(x = avgfacsal, y = netcost)) +
  geom_text(aes(label = type, color = type),
    show.legend = FALSE
  )
```

---

## Two variables - show distribution

- `geom_bin2d()`: bin into rectangles and count
- `geom_density2d()`: smoothed 2d density estimate
- `geom_hex()`: bin into hexagons and count

---

## `geom_hex()`

``` r
library(hexbin)
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_hex()
```

---

## `geom_hex()` and warnings

- Requires installing the [**hexbin**](https://cran.r-project.org/web/packages/hexbin/index.html) package separately!

``` r
install.packages("hexbin")
```

- Otherwise you might see

```
Warning: Computation failed in `stat_binhex()`
```

---

## Two variables

- At least one discrete
  - `geom_count()`: count number of point at distinct locations
  - `geom_jitter()`: randomly jitter overlapping points

- One continuous, one discrete
  - `geom_col()`: a bar chart of pre-computed summaries
  - `geom_boxplot()`: boxplots
  - `geom_violin()`: show density of values in each group

---

## `geom_jitter()`

``` r
ggplot(scorecard, aes(x = type, y = netcost)) +
  geom_point()
```

<img src="index_files/figure-html/unnamed-chunk-46-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[Plot B]

``` r
ggplot(scorecard, aes(x = type, y = netcost)) +
  geom_jitter()
```

<img src="index_files/figure-html/unnamed-chunk-47-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[Plot C]

``` r
ggplot(scorecard, aes(x = type, y = netcost)) +
  geom_jitter()
```

<img src="index_files/figure-html/unnamed-chunk-48-1.png" width="60%" style="display: block; margin: auto;" />
]
]

<countdown-timer class="countdown" id="timer_8983585b" minutes="3" seconds="0" update-every="1" tabindex="0" style="right:0;bottom:0;"></countdown-timer>

---

## `geom_jitter()` and `set.seed()`

``` r
set.seed(123)
ggplot(scorecard, aes(x = type, y = netcost)) +
  geom_jitter()
```

<img src="index_files/figure-html/unnamed-chunk-50-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[Plot B]

``` r
set.seed(123)
ggplot(scorecard, aes(x = type, y = netcost)) +
  geom_jitter()
```

<img src="index_files/figure-html/unnamed-chunk-51-1.png" width="60%" style="display: block; margin: auto;" />
]
.panel[.panel-name[Plot C]

``` r
set.seed(234)
ggplot(scorecard, aes(x = type, y = netcost)) +
  geom_jitter()
```

<img src="index_files/figure-html/unnamed-chunk-52-1.png" width="60%" style="display: block; margin: auto;" />
]
]

---

## Two variables

- One time, one continuous
  - `geom_area()`: area plot
  - `geom_line()`: line plot
  - `geom_step()`: step plot

- Display uncertainty:
  - `geom_crossbar()`: vertical bar with center
  - `geom_errorbar()`: error bars
  - `geom_linerange()`: vertical line
  - `geom_pointrange()`: vertical line with center

- Spatial
  - `geom_sf()`: for map data (more on this later...)

---

## Average net cost per Pell grant recipient

``` r
mean_netcost_pctpell <- scorecard %>%
  mutate(pctpell = round(pctpell, digits = 2)) %>%
  group_by(pctpell) %>%
  summarise(
    n = n(),
    mean_netcost = mean(netcost),
    sd_netcost = sd(netcost)
  ) %>%
  drop_na(mean_netcost)

mean_netcost_pctpell %>% head(6)
```

```
## # A tibble: 6 × 4
##   pctpell     n mean_netcost sd_netcost
##     <dbl> <int>        <dbl>      <dbl>
## 1    0.01     1        2640         NA 
## 2    0.07     3       17447.     12883.
## 3    0.08     3       16578      11709.
## 4    0.1      8       32568.     10281.
## 5    0.12    11       31959       8336.
## 6    0.13    14       29326.      9716.
```

---

## `geom_line()`

``` r
ggplot(mean_netcost_pctpell, aes(x = pctpell, y = mean_netcost)) +
  geom_line()
```

---

## `geom_area()`

``` r
ggplot(mean_netcost_pctpell, aes(x = pctpell, y = mean_netcost)) +
  geom_area()
```

---

## `geom_step()`

``` r
ggplot(mean_netcost_pctpell, aes(x = pctpell, y = mean_netcost)) +
  geom_step()
```

---

## `geom_errorbar()`

.task[
Describe how this plot is constructed and what the points and the lines (error bars) correspond to.
]

``` r
ggplot(mean_netcost_pctpell, aes(x = pctpell, y = mean_netcost)) +
  geom_point() +
  geom_errorbar(aes(
    ymin = mean_netcost - sd_netcost,
    ymax = mean_netcost + sd_netcost
  ))
```
]
.panel[.panel-name[Plot]
<img src="index_files/figure-html/unnamed-chunk-57-1.png" width="60%" style="display: block; margin: auto;" />
]
]

<countdown-timer class="countdown" id="timer_659f1e44" minutes="3" seconds="0" update-every="1" tabindex="0" style="right:0;bottom:0;"></countdown-timer>

---
class: center, inverse, middle

## Let's clean things up a bit!

Meet your new best friend, the [**scales**](https://scales.r-lib.org/) package!

``` r
library(scales)
```

---

# Scales

We need one mapping from our aesthetic layer to our mapping function (e.g. aes). Scales can help with that transformation (use of guides)!

https://scales.r-lib.org/
---

## Let's clean things up a bit!

``` r
library(tidyverse)
library(c3s2datasets)
ggplot(scorecard, aes(x = avgfacsal, y = netcost)) +
  geom_point(alpha = 0.4, size = 2, color = "#012169") +
  scale_x_continuous(labels = label_number(big.mark = ",")) +
  scale_y_continuous(labels = label_dollar()) +
  labs(
    x = "Average faculty salary (USD)",
    y = "Net cost of attendance (USD)",
    title = "Faculty salaries and net cost of attendance in US universities",
    subtitle = "2018-19",
    caption = "Source: College Scorecard"
  )
```

]

]
]

---

# geoms

---

## Three variables

- `geom_tile()`: tile the plane with rectangles
- `geom_raster()`: fast version of `geom_tile()` for equal sized tiles

---

## `geom_tile()`

``` r
ggplot(scorecard, aes(x = type, y = locale)) +
 geom_tile(aes(fill = netcost))
```

---

## Another look at smooth-ish surface

``` r
mean_netcost_type_locale <- scorecard %>%
  group_by(type, locale) %>%
  summarize(mean_netcost = mean(netcost, na.rm = TRUE), .groups = "drop")

mean_netcost_type_locale %>% head(10)
```

```
## # A tibble: 10 × 3
##    type                locale mean_netcost
##    <fct>               <fct>         <dbl>
##  1 Public              City         13428.
##  2 Public              Suburb       14726.
##  3 Public              Town         14433.
##  4 Public              Rural        13040.
##  5 Public              <NA>          3567 
##  6 Private, nonprofit  City         23443.
##  7 Private, nonprofit  Suburb       22024.
##  8 Private, nonprofit  Town         21581.
##  9 Private, nonprofit  Rural        22389.
## 10 Private, for-profit City         28311.
```
]
]
.panel[.panel-name[Plot]

``` r
ggplot(mean_netcost_type_locale, aes(x = type, y = locale)) +
  geom_point(aes(color = mean_netcost), size = 30, pch = "square")
```

<img src="index_files/figure-html/unnamed-chunk-63-1.png" width="60%" style="display: block; margin: auto;" />
]
]

---

## Activity: Pick a geom

1. Display how the value of variable has changed over time
1. Show the detailed distribution of a single continuous variable
1. Focus attention on the overall relationship between two variables in a large dataset
1. Label outlying points in a single variable
]

<countdown-timer class="countdown" id="timer_f107ff79" minutes="3" seconds="0" update-every="1" tabindex="0" style="right:0;bottom:0;"></countdown-timer>

---

# Assignment 2: Refining plots

---

## Assignment 2

.task[
For this assignment, you will submit **SIX plots**: a ‘rough draft and a ‘final’ version for each. ]

https://macs40700.netlify.app/assignments/assign2/ 
--

### Info

* You can use any data for this assignment but the dataset should be the same for both and there should be some connection between the two plots – some narrative of some sort.

* This assignment should be able to stand alone, meaning that all information needed to understand and interpret the plots is provided by you in your writeup. 
* Keep it to 500-750 words
* The final plots should be professional quality–something we could expect to see in a final report, thesis, etc.

---
## A2: Refining components

For each data type (continuous-y and categorical), you will provide three plots:

1. The final plot. This is the plot that you selected to represent the data most effectively.
2. The rough draft / initial plot. This is the initial version of your final plot with all ‘defaults’ – e.g. not at all customized.
3. The alternative plot. This is a second plot option you tried before committing to the final form. For example, maybe you were debating between histograms, dot plots, and box plots. Include one version you tried.

---

# Recap

* Geoms: consider the layering options
* Jitter: consider different seeds
* Scales: way to format axes nicely
* Work on graph ideas for A2