36-315-Lec2

class: inverse, center, middle

# 36-315: Statistical Graphics and Visualization
## Lecture 2

Meghan Hall Department of Statistics & Data Science Carnegie Mellon University May 24, 2021

---
layout: true
<div class="my-footer">cmu-36315.netlify.app</div>
---

# From last time

.large[Syllabus overview]
 
.medium[Email me with any questions]

.large[Why do we visualize data?]
 
.medium[Explore, diagnose, explain]

---

# Updates

.large[Alternate lab]
 
.medium[8:00pm-9:20pm EDT, Tuesday/Thursday]

.large[Office hours]
 
.medium[Listed on Canvas, start Friday!]

---

# Today

.large[The grammar of graphics]
 
.medium[How graphics are constructed in R]

.large[Tidyverse principles]
 
.medium[For any necessary data manipulation]

---

# ggplot

.large[What exactly is the **g**rammar of **g**raphics?]
 
 
.medium["the whole system and structure of a language"]

---

# ggplot

.large[A graphic...]
 
 
.medium[maps the **data** +]
 
 
.medium[to the **aesthetic attributes** +]
 
 
.medium[of **geometric points** +]
 
 
--
.medium[*with possible **statistical transformations** +*]
 
 
.medium[*different **coordinate systems** +*]
 
 
.medium[*and **faceting** *]

---

# ggplot

.huge.center[Data **+** Mapping]

---

# How do we "map" data?

.medium[Encoding data into visual cues dictates how we signify *changes* and *comparisons*]
 
 
.medium[A non-exhaustive list:]

* length
* color (or saturation)
* position
* size
* shape
* area
* volume
* angle

---

class: center

# .left[Today's data]

.pull-left[

]

.pull-right[

]

.right[*Artwork by @allison_horst*]

---

# How are we encoding this?

---

# How are we encoding this?

---

# How are we encoding this?

.right.large[🤔]

---

# How are we encoding this?

.right.large[🤔]

---

# How are we encoding this?

.right.large[🤔]

---

# ggplot

.huge.center[Data **+** Mapping]

1. **layer**
 
 
2. scale
 
 
3. coord
 
 
4. facet
 
 
5. theme

---

# Mapping components: layer

.large[**geom**]
 
.medium[Geometric elements (bars, lines, points, etc.)]

.large[**stat**]
 
.medium[Statistical transformations (summarizing data)]

---

# Mapping components: layer

```r
penguins %>% 
  ggplot(aes(x = species))
```

---

# Mapping components: layer

```r
penguins %>% 
  ggplot(aes(x = species)) +
  geom_bar()
```

---

# Mapping components: layer

```r
penguins %>% 
* group_by(species) %>%
* summarize(mass_mean = mean(body_mass_g, na.rm = TRUE)) %>%
  ggplot(aes(x = species, y = mass_mean)) +
  # what happens if you don't include the stat argument?
  geom_bar(stat = "identity")
```

---

# Mapping components: layer

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point()
```

---

# Mapping components: layer

```r
penguins %>% 
  ggplot(aes(x = sex, y = body_mass_g)) +
  geom_boxplot()
```

---

# Mapping components: layer

```r
penguins %>% 
  ggplot(aes(x = body_mass_g)) +
  geom_histogram()
```

---

# ggplot

.huge.center[Data **+** Mapping]

1. layer
 
 
2. **scale**
 
 
3. coord
 
 
4. facet
 
 
5. theme

---

# Mapping components: scale

.large[data ➡️ aesthetics]
 
.medium[shape, color, size, etc.]

.large[how you interpret the plot]
 
.medium[scales and legends]

---

# Mapping components: scale

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point()
```

---

# Mapping components: scale

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point(color = "blue", size = 2)
```

---

# Mapping components: scale

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point(aes(color = species), size = 2)
```

---

# Mapping components: scale

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point(aes(color = species), size = 2) +
  scale_color_manual(values = c("#aa6600","#666666","#224477"))
```

---

# Mapping components: scale

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point(aes(color = species), size = 2) +
  scale_color_manual(values = c("#aa6600","#666666","#224477")) +
  scale_x_continuous(name = "Bill Length (mm)", 
                     breaks = seq(30, 60, by = 5), limits = c(30, 60))
```

---

# ggplot

.huge.center[Data **+** Mapping]

1. layer
 
 
2. scale
 
 
3. **coord**
 
 
4. facet
 
 
5. theme

---

# Mapping components: coord

.large[x and y]
 
.medium[or latitude and longitude, or radius and angle]

.large[we'll discuss more about maps later!]

---

# Mapping components: coord

```r
penguins %>% 
  group_by(species) %>%
  summarize(mass_mean = mean(body_mass_g, na.rm = TRUE)) %>% 
  ggplot(aes(x = species, y = mass_mean)) +
  geom_bar(stat = "identity")
```

---

# Mapping components: coord

```r
penguins %>% 
  group_by(species) %>%
  summarize(mass_mean = mean(body_mass_g, na.rm = TRUE)) %>% 
  ggplot(aes(x = species, y = mass_mean)) +
  geom_bar(stat = "identity") +
* coord_flip()
```

---

# Mapping components: coord

```r
penguins %>% 
  count(sex) %>% 
  ggplot(aes(x = "", y = n, fill = sex)) +
  geom_bar(stat = "identity", width = 1, color = "white") +
* coord_polar("y", start = 0) +
  theme_void()
```

.right.large[🤔]

---

# ggplot

.huge.center[Data **+** Mapping]

1. layer
 
 
2. scale
 
 
3. coord
 
 
4. **facet**
 
 
5. theme

---

# Mapping components: facet

.large[create small multiples]
 
.medium[useful for the *explore* part of data viz]

---

# Mapping components: facet

```r
penguins %>% 
  filter(!is.na(sex)) %>% 
  ggplot(aes(x = sex, y = body_mass_g)) +
  geom_point(position = "jitter")
```

---

# Mapping components: facet

```r
penguins %>% 
  filter(!is.na(sex)) %>% 
  ggplot(aes(x = sex, y = body_mass_g)) +
  geom_point(position = "jitter") +
* facet_wrap(~species)
```

---

# ggplot

.huge.center[Data **+** Mapping]

1. layer
 
 
2. scale
 
 
3. coord
 
 
4. facet
 
 
5. **theme**

---

# Mapping components: theme

.large[adjust individual pieces of the plot]
 
.medium[font size, gridlines, legend position, etc.]

.large[or go full out with a custom theme!]

---

# Mapping components: theme

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point(aes(color = species), size = 2) +
  scale_color_manual(values = c("#aa6600","#666666","#224477"))
```

---

# Mapping components: theme

```r
penguins %>% 
  ggplot(aes(x = bill_length_mm, y = bill_depth_mm)) +
  geom_point(aes(color = species), size = 2) +
  scale_color_manual(values = c("#aa6600","#666666","#224477")) +
  theme(legend.position = "top", legend.title = element_blank())
```

---

# Mapping components: theme

```r
penguins %>% 
  filter(!is.na(sex)) %>% 
  ggplot(aes(x = sex, y = body_mass_g)) +
  geom_point(position = "jitter") +
  facet_wrap(~species)
```

---

# Mapping components: theme

```r
penguins %>% 
  filter(!is.na(sex)) %>% 
  ggplot(aes(x = sex, y = body_mass_g)) +
  geom_point(position = "jitter") +
  facet_wrap(~species) +
  theme(panel.grid.major.x = element_blank())
```

---

# ggplot in review

.large[A graphic...]
 
 
.medium[maps the **data** +]
 
 
.medium[to the **aesthetic attributes** +]
 
 
.medium[of **geometric points** +]
 
 
.medium[*with possible **statistical transformations** +*]
 
 
.medium[*different **coordinate systems** +*]
 
 
.medium[*and **faceting** *]

---

# What's the tidyverse?

.center[![tidyverse packages](figs/Lec2/tidyverse.png)]

---

# What's the tidyverse?

.large[An "opinionated" set of packages]
 
.medium[Similar philosophies, grammar, data structures]

.large[`dplyr`, `stringr`, `tidyr`, `readr`, `forcats`]
 
.medium[Useful for basic data manipulation]

.large[Best resource: [r4ds.had.co.nz/](https://r4ds.had.co.nz/)]

---