Mapping data to graphics, amounts, and proportions

class: center middle main-title section-title-4

# Mapping data to graphics, amounts, and proportions

.class-info[

**Week 7**

AEM 2850 / 5850 : R for Business Analytics<br>
Cornell Dyson<br>
Fall 2025

Acknowledgements: 
[Andrew Heiss](https://datavizm20.classes.andrewheiss.com), 
[Claus Wilke](https://wilkelab.org/SDS375/)

]

---

# Announcements

This week marks the start of the data visualization component of this course

No homework this week due to Fall Break 😎

**Group project due Friday, Nov. 7 at 11:59pm** (full details to come soon!)

**Group project groups due on canvas this Friday (October 10)**
  - Form groups of 4
  - All students in each group must be enrolled at the same course level
  - If unable to form a group of 4, please state that in your submission

We are grading Prelim 1 and will update you when we can

Questions before we get started?

---

# Plan for this week

.pull-left[
### Tuesday
  - [Course progress](#progress)
  - [Prologue](#prologue)
  - [Data, aesthetics, <br> & the grammar of graphics](#grammar-of-graphics)
  - [Amounts](#amounts)
  - [Plotting amounts using ggplot](#plotting-amounts)
  - [example-07-1](#example-1)
  - [Reference: More examples](#more-examples)
]

.pull-right[
### Thursday
  - [Proportions](#proportions)
  - [example-07-2](#example-2)
  - [Reference: Additional layers](#additional-layers)
]

---
class: inverse, center, middle
name: progress

# Course progress

---

# Course objectives reminder

1. Develop basic proficiency in `R` programming
2. Understand data structures and manipulation
3. Describe effective techniques for data visualization and communication
4. Construct effective data visualizations
5. Utilize course concepts and tools for business applications

---

# Where we've been

1. **Develop basic proficiency in `R` programming**
2. **Understand data structures and manipulation**
3. Describe effective techniques for data visualization and communication
4. Construct effective data visualizations
5. Utilize course concepts and tools for business applications

---

# Where we're going next

1. Develop basic proficiency in `R` programming
2. Understand data structures and manipulation
3. **Describe effective techniques for data visualization and communication**
4. **Construct effective data visualizations**
5. Utilize course concepts and tools for business applications

---

# Schedule overview

#### Weeks 1-5: Programming Foundations

#### Weeks 7-10: Data Visualization Foundations

#### Weeks 11+: Special Topics (mix of programming and dataviz)

See [aem2850.toddgerarden.com/schedule](https://aem2850.toddgerarden.com/schedule/) for details

---
class: inverse, center, middle
name: prologue

# Prologue

---

# Remember our concentrations?

How might we visualize these amounts?

.less-left[

```
## # A tibble: 12 × 2
##    concentration count
##    <chr>         <int>
##  1 fin              39
##  2 ba               20
##  3 food             10
##  4 strat            10
##  5 mktg              9
##  6 econ              7
##  7 acct              6
##  8 entr              6
##  9 dev               5
## 10 sus               3
## 11 mgmt              3
## 12 oth               3
```
]

.more-right[
<img src="07-slides_files/figure-html/unnamed-chunk-2-1.png" width="1008" style="display: block; margin: auto;" />
]

---

# What are our favorite public companies?

---
class: inverse, center, middle
name: grammar-of-graphics

# Data, aesthetics,<br>& the grammar of graphics

---

# Mapping data to aesthetics

.pull-left.center[

]

.pull-right[

### Data

A column in a dataset

### Aesthetics

Visual properties of a graph

Position, shape, color, etc.

]

---

# An example: Health vs wealth

.pull-left[
<img src="img/07/gapminder-screenshot.png" width="100%" style="display: block; margin: auto;" />
]

.pull-right[
<table>
  <tr>
    <th class="cell-left">Data</th>
    <th class="cell-left">Aesthetic</th>
    <th class="cell-left">Geometry</th>
  </tr>
  <tr>
    <td class="cell-left">Wealth</td>
    <td class="cell-left">Position (x)</td>
    <td class="cell-left">Point</td>
  </tr>
  <tr>
    <td class="cell-left">Health</td>
    <td class="cell-left">Position (y)</td>
    <td class="cell-left">Point</td>
  </tr>
  <tr>
    <td class="cell-left">Continent</td>
    <td class="cell-left">Color</td>
    <td class="cell-left">Point</td>
  </tr>
  <tr>
    <td class="cell-left">Population</td>
    <td class="cell-left">Size</td>
    <td class="cell-left">Point</td>
  </tr>
</table>
]

---

# Barebones `ggplot2::ggplot()` template

``` r
library(tidyverse) # tidyverse loads the package ggplot2
```

We need to specify data, aesthetic mapping, and geometry:

``` r
ggplot(
  data = DATA,
  mapping = aes(AESTHETIC MAPPINGS)
) +
  GEOM_FUNCTION()
```

Or, in the context of a data wrangling pipeline:

``` r
DATA |> 
  ... |> # intermediate data wrangling (optional)
  ggplot(aes(AESTHETIC MAPPINGS)) +
  GEOM_FUNCTION()
```

---

# Mapping from `gapminder` to aesthetics

<table>
 <thead>
  <tr>
   <th style="text-align:center;"> country </th>
   <th style="text-align:center;"> continent </th>
   <th style="text-align:center;"> gdpPercap </th>
   <th style="text-align:center;"> lifeExp </th>
   <th style="text-align:center;"> pop </th>
  </tr>
 </thead>
<tbody>
  <tr>
   <td style="text-align:center;"> Afghanistan </td>
   <td style="text-align:center;"> Asia </td>
   <td style="text-align:center;"> 974.5803384 </td>
   <td style="text-align:center;"> 43.828 </td>
   <td style="text-align:center;"> 31889923 </td>
  </tr>
  <tr>
   <td style="text-align:center;"> Albania </td>
   <td style="text-align:center;"> Europe </td>
   <td style="text-align:center;"> 5937.029526 </td>
   <td style="text-align:center;"> 76.423 </td>
   <td style="text-align:center;"> 3600523 </td>
  </tr>
  <tr>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
  </tr>
</tbody>
</table>

``` r
*_________ |>
  ggplot(aes(x = _________,
             y = _______,
             color = _________,
             size = ___)) +
  geom______() +
  scale_x_log10() + theme_classic(base_size = 20) # ignore this line for now
```

Let's fill in the blanks... how do we start?

---

# Mapping from `gapminder` to aesthetics
<table>
 <thead>
  <tr>
   <th style="text-align:center;"> country </th>
   <th style="text-align:center;"> continent </th>
   <th style="text-align:center;"> gdpPercap </th>
   <th style="text-align:center;"> lifeExp </th>
   <th style="text-align:center;"> pop </th>
  </tr>
 </thead>
<tbody>
  <tr>
   <td style="text-align:center;"> Afghanistan </td>
   <td style="text-align:center;"> Asia </td>
   <td style="text-align:center;"> 974.5803384 </td>
   <td style="text-align:center;"> 43.828 </td>
   <td style="text-align:center;"> 31889923 </td>
  </tr>
  <tr>
   <td style="text-align:center;"> Albania </td>
   <td style="text-align:center;"> Europe </td>
   <td style="text-align:center;"> 5937.029526 </td>
   <td style="text-align:center;"> 76.423 </td>
   <td style="text-align:center;"> 3600523 </td>
  </tr>
  <tr>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
   <td style="text-align:center;"> … </td>
  </tr>
</tbody>
</table>

``` r
gapminder |> 
* ggplot(aes(x = _________,
*            y = _______,
             color = _________,
             size = ___)) +
  geom______() +
  scale_x_log10() + theme_classic(base_size = 20) # ignore this line for now
```

Let's fill in the blanks... what should we map to x? y?

---

``` r
gapminder |> 
  ggplot(aes(x = gdpPercap,
             y = lifeExp,
*            color = _________,
*            size = ___)) +
  geom______() +
  scale_x_log10() + theme_classic(base_size = 20) # ignore this line for now
```

Let's fill in the blanks... what should we map to color? size?

---

``` r
gapminder |> 
  ggplot(aes(x = gdpPercap,
             y = lifeExp,
             color = continent,
             size = pop)) +
* geom______() +
  scale_x_log10() + theme_classic(base_size = 20) # ignore this line for now
```

Let's fill in the blanks... what geometry should we use?

---

``` r
gapminder |> 
  ggplot(aes(x = gdpPercap,
             y = lifeExp,
             color = continent,
             size = pop)) +
  geom_point() +
  scale_x_log10() + theme_classic(base_size = 20) # ignore this line for now
```

All done! Let's see what we get...

---

# Health and wealth

---

# Grammar components as layers

.pull-left[
So far we know about data, aesthetics, and geometries

Think of these components as **layers**

Add to foundational `ggplot()` with `+`

Why `+` and not `|>`?
> ggplot2 was written before the pipe was discovered

Treat the `+` the same as `|>`

]

.pull-right[

&nbsp;

![](img/07/ggplot-layers-short@4x.png)
]

???

Layer analogy borrowed from [Thomas Lin Pedersen](https://www.data-imaginist.com/) and his ["Drawing Anything with ggplot2" workshop](https://github.com/thomasp85/ggplot2_workshop).

Source for quote: [R4DS (2e)](https://r4ds.hadley.nz/workflow-style.html#ggplot2)

---

# Additional layers

.pull-left[
There are many other grammatical layers we can use to describe graphs!

We can sequentially add layers to the foundational `ggplot()` plot to create complex figures

We will primarily learn by doing, though this slide deck contains [a preview of additional layers](#additional-layers) in case you need some bedtime reading
]

.pull-right[
![](img/07/ggplot-layers@4x.png)
]

---
class: inverse, center, middle
name: amounts

# Amounts

---

# Yay bar plots!

We are a lot better at visualizing line lengths than angles and areas

.pull-left[

]

.pull-right[

]

---

# Oh no bar plots!

.center[
<figure>
<img src="img/07/obamacareenrollment-fncchart.jpg" alt="Fox News Obamacare enrollment" title="Fox News Obamacare enrollment" width="85%">
</figure>
]

---

# What went wrong?

.center[
<figure>
<img src="img/07/obamacareenrollment-fncchart.jpg" alt="Fox News Obamacare enrollment" title="Fox News Obamacare enrollment" width="85%">
</figure>
]

---

# What went wrong?

.pull-left.center[
<figure>
<img src="img/07/female-height.png" alt="Average female height per country" title="Average female height per country" width="100%">
</figure>
]

.pull-right[

**At least two problems:**

1. truncated y axis
2. area scales faster than height

.small[This terrible figure brought to you by a former AEM 2850 / 5850 student!]
]

---

# General rules for bar charts

Useful when the length of the bar is all that matters

Bar charts should always start at zero
- Or: don't use bars!

Don't use bars for summary statistics. You throw away too much information.
- We will come back to visualizing distributions / uncertainty next week

---
class: inverse, center, middle
name: plotting-amounts

# Plotting amounts using ggplot

---

# Plotting amounts using ggplot

We'll use a summarized version of the gapminder dataset for examples

``` r
library(gapminder)
gapminder_continents <- gapminder |>
  filter(year == 2007) |>  # only look at 2007
  count(continent) |>      # count the number of countries per continent
  arrange(desc(n))         # sort by count, descending
gapminder_continents
```

```
## # A tibble: 5 × 2
##   continent     n
##   <fct>     <int>
## 1 Africa       52
## 2 Asia         33
## 3 Europe       30
## 4 Americas     25
## 5 Oceania       2
```

---

# Start with a simple bar plot

.left-code.small-code[

``` r
gapminder_continents |> 
  ggplot(aes(x = continent, # continent to x
             y = n)) +      # n countries to y
* geom_col()                # add bars
```
How could we improve this?
]

.right-plot[

![](07-slides_files/figure-html/gapminder-bars-1.png)

]

---

# Start with a simple bar plot

.left-code.small-code[

``` r
gapminder_continents |> 
  ggplot(aes(x = continent, # continent to x
             y = n)) +      # n countries to y
* geom_col()                # add bars
```
Is "n" a good axis title?

Do we need "continent" at all?
]

.right-plot[

![](07-slides_files/figure-html/gapminder-bars-2-1.png)

]

---

# Add some labels

.left-code.small-code[

``` r
gapminder_continents |> 
  ggplot(aes(x = continent, 
             y = n)) +
  geom_col() +
* labs(x = NULL, y = "Number of countries")
```

Is alphabetical the best ordering?
]

.right-plot[

![](07-slides_files/figure-html/gapminder-bars-labeled-1.png)

]

---

# Order by data value

.left-code.small-code[

``` r
gapminder_continents |> 
* ggplot(aes(x = fct_reorder(continent, n),
             y = n)) +
  geom_col() +
  labs(x = NULL, y = "Number of countries")
```

`fct_reorder(continent, n)` means "reorder the factor variable continent by n, smallest to largest"

Factor is R's term for variables that have a fixed and known set of possible values
]

.right-plot[

![](07-slides_files/figure-html/gapminder-bars-ordered-1.png)

]

---

# Order by data value, descending

.left-code.small-code[

``` r
gapminder_continents |> 
* ggplot(aes(x = fct_reorder(continent, -n),
             y = n)) +
  geom_col() +
  labs(x = NULL, y = "Number of countries")
```

`fct_reorder(continent, -n)` means "reorder the factor variable continent by n, largest to smallest"
]

.right-plot[

![](07-slides_files/figure-html/gapminder-bars-reversed-1.png)

]

---

# Higher dimensional datasets

`as_factor()` treats `year` as distinct categories, not a continuous measure

.left-code[

``` r
gapminder |>
  filter(year==1957 | year==2007) |> 
  group_by(continent, year) |> 
  summarize(
    lifeExp = sum(lifeExp * pop) / sum(pop)
  ) |> 
  ggplot(aes(x = continent,
             y = lifeExp,
*            fill = as_factor(year))) +
  geom_col() +
  labs(x = NULL, 
       y = "Life expectancy", fill = "Year")
```
What is wrong with this visualization of life expectancy in two different years?
]

.right-plot[

![](07-slides_files/figure-html/geom_bar-ungrouped-1.png)

]

---

# Grouped bar charts

Use grouped bars or facets for higher dimensional datasets

.left-code[

``` r
gapminder |>
  filter(year==1957 | year==2007) |> 
  group_by(continent, year) |> 
  summarize(
    lifeExp = sum(lifeExp * pop) / sum(pop)
  ) |> 
  ggplot(aes(x = continent,
             y = lifeExp,
             fill = as_factor(year))) + 
* geom_col(position = "dodge") +
  labs(x = NULL, 
       y = "Life expectancy", fill = "Year")
```
]

.right-plot[

![](07-slides_files/figure-html/geom_bar-grouped-1.png)

]

---

# Facets

Use grouped bars or facets for higher dimensional datasets

.left-code[

``` r
gapminder |>
  filter(year==1957 | year==2007) |> 
  group_by(continent, year) |> 
  summarize(
    lifeExp = sum(lifeExp * pop) / sum(pop)
  ) |> 
  ggplot(aes(x = continent,
             y = lifeExp,
             fill = as_factor(year))) + 
* geom_col() +
* facet_wrap(vars(year)) +
* guides(fill = "none") +
  labs(x = NULL, 
       y = "Life expectancy")
```
]

.right-plot[

![](07-slides_files/figure-html/geom_bar-faceted-1.png)

]

---

# Alternative: Dots instead of bars

Dots are preferable if we want to truncate the axes

.left-code[

``` r
gapminder |>
  filter(year == 2007, continent == "Americas") |>
  ggplot(aes(x = lifeExp,
             y = fct_reorder(country, lifeExp))) +
* geom_point() +
  guides(color = "none") +
  labs(x = "Life expectancy (years)", y = NULL) +
  theme_minimal()
```
]

.right-plot[

![](07-slides_files/figure-html/dots-1.png)

]

---
class: inverse, center, middle
name: example-1

# example-07-1:<br>amounts-practice.R

---
class: inverse, center, middle
name: more-examples

# Reference: More examples

---

# Wait, what about geom_bar?

Use `geom_bar` to count and plot in one step

.left-code[

``` r
*gapminder |> # old data: rows are countries
  filter(year == 2007) |>
  ggplot(aes(x = continent)) + # note: no y arg
* geom_bar() +
  labs(x = NULL, y = "Number of countries")
```
]

.right-plot[

![](07-slides_files/figure-html/geom_bar-1.png)

]

---

# Wait, what about geom_bar?

Here we can reorder by frequency using `fct_infreq`

.left-code[

``` r
gapminder |>
  filter(year == 2007) |>
* ggplot(aes(x = fct_infreq(continent))) +
  geom_bar() +
  labs(x = NULL, y = "Number of countries")
```
]

.right-plot[

![](07-slides_files/figure-html/geom_bar-ordered-1.png)

]

---

# We can also flip geom_col/bar axes

Simply use `y = ` instead of `x = ` for the aesthetic mapping

.left-code[

``` r
gapminder |>
  filter(year == 2007) |>
* ggplot(aes(y = fct_infreq(continent))) +
  geom_bar() +
* labs(x = "Number of countries", y = NULL)
```
]

.right-plot[

![](07-slides_files/figure-html/geom_bar-y-1.png)

]

---

# We can also flip geom_col/bar axes

`fct_rev()` reverses the order of factors

.left-code[

``` r
gapminder |>
  filter(year == 2007) |>
* ggplot(aes(y = fct_rev(fct_infreq(continent)))) +
  geom_bar() +
  labs(x = "Number of countries", y = NULL)
```
]

.right-plot[

![](07-slides_files/figure-html/geom_bar-y-rev-1.png)

]

---
class: inverse, center, middle
name: proportions

# Proportions

---

# Last class we plotted amounts

---

# How else could we visualize these data?

---

# Have you done any programming before?

.pull-left-3[
<img src="07-slides_files/figure-html/unnamed-chunk-8-1.png" width="648" style="display: block; margin: auto;" />
]

.pull-middle-3[
<img src="07-slides_files/figure-html/unnamed-chunk-9-1.png" width="648" style="display: block; margin: auto;" />
]

.pull-right-3[
<img src="07-slides_files/figure-html/unnamed-chunk-10-1.png" width="648" style="display: block; margin: auto;" />
]

Which do you think is best?

Does it depend on what you want to communicate?

---