03_dataviz.Rmd

---
title: "Data visualization in R"
output: 
  learnr::tutorial:
    progressive: true   
    allow_skip: true
runtime: shiny_prerendered
---

```{r setup, include=FALSE}
library(learnr)
library(tidyverse)
library(viridis)
library(ggrepel)
library(scales)
library(lubridate)

vac <- read_csv("https://www.dropbox.com/s/3c7kuqja1pv76r8/country_vaccinations.csv?dl=1")

vac_top <- vac %>%
    group_by(country) %>%
    filter(date == max(date)) %>%
    ungroup() %>%
    arrange(desc(total_vaccinations_per_hundred)) %>%
    slice(1:10)

countries <- c("Israel", "Switzerland", "Norway", "Denmark", "Belgium")
vac_subset <- vac %>%
    filter(country %in% countries)

vac_subset_last <- vac_subset %>%
    group_by(country) %>%
    filter(min_rank(desc(date)) == 1)

shapes <- 21:25
names(shapes) <- countries
```

# A grammar for graphics

## Welcome!

### Today's programme

We will explore how to build informative and beautiful visualizations using the [ggpplot2](https://ggplot2.tidyverse.org/index.html) package, part of the [tidyverse](https://www.tidyverse.org/) suite. 

In the interactive tutorial part, we will cover the following topics

* ggplot2 basics
* Aesthetics
* Geometric shapes
* Facets
* Scales
* Annotations
* Themes

Throughout we will progressively build up an example visualization of COVID19 vaccination progress for selected countries:

![*COVID19 vaccination rates over time*](https://www.dropbox.com/s/7bxebcgwgpdxax0/plot.png?dl=1){width=80%}

Following the interactive tutorial, there will be individual group exercises, where you will apply the learned techniques to build visualizations exploring different aspects of the example dataset.


**Some useful resources**

[ggplot2 cheat sheet](https://raw.githubusercontent.com/rstudio/cheatsheets/main/data-visualization.pdf)  
R for Data Science, chapter [visualization](https://r4ds.had.co.nz/data-visualisation.html)  
[Fundamentals of Data Visualization](https://clauswilke.com/dataviz/)  

### Interactive tutorial format
This is what a code exercise block looks like

```{r intro, exercise = TRUE}
hist(rnorm(1000))
```

In tutorial sections, example code is provided by clicking the 'Solution' button
```{r intro1, exercise=TRUE}
hist(rnorm(1000))
```

```{r intro1-solution}
hist(rnorm(1000), breaks=100)
```

In the exercise sections, hints can be progressively revealed by clicking the 'Hints' button

```{r intro2, exercise = TRUE}
hist(rnorm(1000))
```

```{r intro2-hint}
hist(rnorm(1000), breaks=...)
```

```{r intro2-solution}
hist(rnorm(1000), breaks=100)
```
## Basics

### Set up

pre-loading some necessary libraries

```{r, eval=FALSE}
library(tidyverse)
library(viridis)
library(ggrepel)
library(scales)
library(lubridate)
```

we'll use a [COVID19 vaccinations progress](https://www.kaggle.com/gpreda/covid-world-vaccination-progress) dataset for this tutorial


```{r, eval=FALSE}
vac <- read_csv("https://www.dropbox.com/s/3c7kuqja1pv76r8/country_vaccinations.csv?dl=1")

countries <- c("Israel", "Switzerland", "Norway", "Denmark", "Belgium")
vac_subset <- vac %>%
    filter(country %in% countries)
```

### Let's explore the data
```{r explore-data, exercise = TRUE}
vac
```

```{r explore-data-solution}
vac
vac %>%
    head()
vac %>%
    colnames()
```

### ggplot2 syntax
the most basic ggplot2 visualizations are built using the following minimal syntax:

```{r, eval = FALSE}
ggplot(data=<DATA>) + 
  <GEOM_FUNCTION>(mapping = aes(<MAPPINGS>))
```

### Simple plot using 2 variables
```{r coord1, exercise = TRUE}
ggplot()
```
```{r coord1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations))
```

### Modifying coordinate systems
```{r coord2, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations))
```
```{r coord2-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations))

### explicit coordinate system
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) + 
    coord_cartesian()

### zoom in
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) + 
    coord_cartesian(ylim = c(0, 1e6))

### other coordinate system
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) + 
    coord_polar()
```

### Short exercise 1

Plot the rate of people fully vaccinated (variable 'people_fully_vaccinated_per_hundred') against rate of people having received at least one vaccine dose (variable 'people_vaccinated_per_hundred'). 
```{r coord-ex1, exercise = TRUE}
vac_subset
```
```{r coord-ex1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = people_vaccinated_per_hundred, y = people_fully_vaccinated_per_hundred))
```


## Aesthetics

### Mapping aesthetics to variables
```{r aes1, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country)) 
```
```{r aes1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country)) 

### map colour to daily vaccinations
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw)) 

### map size and transparency to daily vaccinations
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations,  size = daily_vaccinations_raw, alpha = daily_vaccinations_raw))
```

### Mapping versus setting an aesthaetic
```{r aes2, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw)) 
```
```{r aes2-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw)) 

ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = "daily_vaccinations_raw")) 
```

### Start building the example visualization

![*COVID19 vaccination rates over time*](https://www.dropbox.com/s/7bxebcgwgpdxax0/plot.png?dl=1){width=80%}

```{r example1, exercise = TRUE}
vac_subset
```
```{r example1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) 
```

## Geometric shapes

### Simple point geom
```{r geom1, exercise =TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) 
```

### Aesthetics of points
```{r geom2, exercise = TRUE}
df <- tibble(x = factor(0:25), y = 1)
```
```{r geom2-solution}
df <- tibble(x = factor(0:25), y = 1)

### geom_point shapes
ggplot(df) +
    geom_point(mapping = aes(x = x, y = y), shape = 0:25, size = 4)

### geom_point colours
ggplot(df) +
    geom_point(mapping = aes(x = x, y = y), shape = 0:25, colour = 1:26, fill = "grey", size = 4, stroke = 1)
```

### Aesthetics of lines
```{r geom3, exercise = TRUE}
df <- tibble(x = 0:5, xend = 0:5, y = 0, yend = 10)
```
```{r geom3-solution}
df <- tibble(x = 0:5, xend = 0:5, y = 0, yend = 10)

### geom_line linetypes
ggplot(df) +
    geom_segment(mapping = aes(x = x, xend = xend, y = y, yend = yend), size = 3, linetype = 1:6)

### path vs line
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations))
ggplot(vac_subset) +
    geom_line(mapping = aes(x = date, y = total_vaccinations))
```

### Examples of other geoms
```{r geom4, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, fill = country)) +
    geom_hline(yintercept = 1e6, linetype = "dashed")
```
```{r geom4-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, fill = country)) +
    geom_hline(yintercept = 1e6, linetype = "dashed")

ggplot(vac_subset) +
    geom_col(mapping = aes(x = date, y = total_vaccinations, fill = country)) 
```

### Positional adjustments
```{r geom5, exercise = TRUE}
ggplot(vac_subset) +
    geom_col(mapping = aes(x = date, y = total_vaccinations, fill = country), position = position_stack()) 
```

```{r geom5-solution}
ggplot(vac_subset) +
    geom_col(mapping = aes(x = date, y = total_vaccinations, fill = country), position = position_stack()) 

### dodge bars
ggplot(vac_subset) +
    geom_col(mapping = aes(x = date, y = total_vaccinations, fill = country), position = position_dodge()) 

### proportions of full
ggplot(vac_subset) +
    geom_col(mapping = aes(x = date, y = total_vaccinations, fill = country), position = position_fill()) 

### jitter
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations), size = 1)
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations), size = 1, position = position_jitter(height = 5e5, width = 0.15), alpha = 0.5)
```

### Use different data set for geoms
```{r geom6, exercise = TRUE}
vac_subset_last <- vac_subset %>%
    group_by(country) %>%
    filter(min_rank(desc(date)) == 1)
vac_subset_last
```
```{r geom6-solution}
vac_subset_last <- vac_subset %>%
    group_by(country) %>%
    filter(min_rank(desc(date)) == 1)
vac_subset_last

ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    geom_point(mapping = aes(x = date, y = total_vaccinations), data = vac_subset_last, size = 3, shape = 17)
```

### Continue building the example visualization

![*COVID19 vaccination rates over time*](https://www.dropbox.com/s/7bxebcgwgpdxax0/plot.png?dl=1){width=80%}

```{r example2, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw))
```
```{r example2-solution}
### plot points only on last vaccination day
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last)
```

### Short exercise 2
Plot the rate of people fully vaccinated (variable 'people_fully_vaccinated_per_hundred') against rate of people having received at least one vaccine dose (variable 'people_vaccinated_per_hundred'). Map colour to the country (variable 'country'), and use both points and lines as geoms

```{r geom-ex1, exercise = TRUE}
vac_subset
```
```{r geom-ex1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = people_vaccinated_per_hundred, y = people_fully_vaccinated_per_hundred, colour = country)) +
    geom_line(mapping = aes(x = people_vaccinated_per_hundred, y = people_fully_vaccinated_per_hundred, colour = country))
```

## Facets

### Simple faceting mapping one variable to panels
```{r facet1, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    facet_wrap(~country)
```
```{r facet1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    facet_wrap(~country)

### panels in grid
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    facet_grid(country ~ .)
```

### Faceting two or more variables
```{r facet2, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    facet_grid(country ~ .)
```
```{r facet2-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    facet_grid(country ~ .)

### 2 variables in grid
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    facet_grid(country ~ vaccines)

### muliple variables per panel
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    facet_wrap(~country+vaccines)
```

### Continue building the example visualization

![*COVID19 vaccination rates over time*](https://www.dropbox.com/s/7bxebcgwgpdxax0/plot.png?dl=1){width=80%}

```{r example3, exercise = TRUE}
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last)
```
```{r example3-solution}
### facet plot by vaccination manufacturers
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last) + 
    facet_grid(vaccines ~ .)
```

## Scales

### Colour scales

```{r scale1, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw))
```
```{r scale1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw))

### two colour gradient
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw)) +
    scale_colour_gradient(low = "blue", high = "red")

### multicolor gradient
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw)) +
    scale_colour_gradientn(colours = rainbow(100))

### discrete colour scale
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = country))
```

### Specialized colour palettes

```{r scale2, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = country)) +
    scale_color_brewer()
```
```{r scale2-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = country)) +
    scale_color_brewer()

### Color brewer palettes
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = country)) +
    scale_color_brewer(palette = "Set1")

ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = country)) +
    scale_color_brewer(palette = "Spectral")

ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw)) +
    scale_colour_distiller(palette = "Spectral")

### Viridis palettes
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw)) +
    scale_colour_viridis()
```

[Color brewer](https://colorbrewer2.org/)  
[Viridis color palettes](https://cran.r-project.org/web/packages/viridis/vignettes/intro-to-viridis.html)


### Manual scale for discrete variables

```{r scale3, exercise = TRUE}
shapes <- 21:25
names(shapes) <- countries
```
```{r scale3-solution}
shapes <- 21:25
names(shapes) <- countries

ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    scale_colour_viridis() +
    scale_shape_manual(values = shapes)
```

### Adjusting scales

```{r scale4, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    scale_colour_viridis() +
    scale_shape_manual(values = shapes) +
    scale_y_continuous(trans = "log10")
```
```{r scale4-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    scale_colour_viridis() +
    scale_shape_manual(values = shapes) +
    scale_y_continuous(trans = "log10")

### set limits
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    scale_colour_viridis() +
    scale_shape_manual(values = shapes) +
    scale_y_continuous(limits = c(0, 1e6))

### custom labels
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country, colour = daily_vaccinations_raw)) +
    scale_colour_viridis() +
    scale_shape_manual(values = shapes) +
    scale_y_continuous(labels = label_number())
```

### Continue building the example visualization

![*COVID19 vaccination rates over time*](https://www.dropbox.com/s/7bxebcgwgpdxax0/plot.png?dl=1){width=80%}

```{r example4, exercise=TRUE}
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last) + 
    facet_grid(vaccines ~ .)
```
```{r example4-solution}
### add custom scales and labels
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last) + 
    facet_grid(vaccines ~ .) + 
    scale_y_continuous(labels = label_comma()) +
    scale_colour_viridis(labels = label_comma()) +
    scale_fill_viridis(labels = label_comma()) +
    scale_shape_manual(values = shapes)
```

## Annotations

### Add text to plot
```{r ann1, exercise = TRUE}
vac_subset_date <- vac_subset %>%
    filter(date > ymd(20210115), date < ymd(20210120))
```
```{r ann1-solution}
vac_subset_date <- vac_subset %>%
    filter(date > ymd(20210115), date < ymd(20210120))

### add text 
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations), colour = "grey") +
    geom_text(mapping = aes(x = date, y = total_vaccinations, label = total_vaccinations), size = 3, data = vac_subset_date)

### avoid overplotting by repelling text 
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations), colour = "grey") +
    geom_text_repel(mapping = aes(x = date, y = total_vaccinations, label = total_vaccinations), size = 2, min.segment.length = 0, data = vac_subset_date, max.overlaps=Inf)

### label format 
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations), colour = "grey") +
    geom_label_repel(mapping = aes(x = date, y = total_vaccinations, label = total_vaccinations), size = 2, min.segment.length = 0, data = vac_subset_date, max.overlaps=Inf)
```

### Continue building the example visualization

![*COVID19 vaccination rates over time*](https://www.dropbox.com/s/7bxebcgwgpdxax0/plot.png?dl=1){width=80%}

```{r example5, exercise = TRUE}
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last) + 
    facet_grid(vaccines ~ .) + 
    scale_y_continuous(labels = label_comma()) +
    scale_colour_viridis(labels = label_comma()) +
    scale_fill_viridis(labels = label_comma()) +
    scale_shape_manual(values = shapes)
```
```{r example5-solution}
### add country label
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last) + 
    geom_text_repel(aes(x = date, y = total_vaccinations, label = country), size = 2, color = "black", direction = "y", data = vac_subset_last) +
    facet_grid(vaccines ~ .) + 
    scale_y_continuous(labels = label_comma()) +
    scale_colour_viridis(labels = label_comma()) +
    scale_fill_viridis(labels = label_comma()) +
    scale_shape_manual(values = shapes)
```

## Themes and polishing plots

### Examples of pre-built themes

```{r theme1, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    facet_grid(vaccines ~ .) +
    theme_gray()
```
```{r theme1-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    facet_grid(vaccines ~ .) +
    theme_gray()

### black & white theme 
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    facet_grid(vaccines ~ .) +
    theme_bw()

### classic theme
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    facet_grid(vaccines ~ .) +
    theme_classic()

### theme without elements
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    facet_grid(vaccines ~ .) +
    theme_void()
```

### Customize themes

```{r theme2, exercise = TRUE}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    facet_grid(vaccines ~ .) +
    theme_bw() +
    theme(panel.grid.major.x = element_blank(), panel.grid.minor.x = element_blank())
```
```{r theme2-solution}
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    facet_grid(vaccines ~ .) +
    theme_bw() +
    theme(panel.grid.major.x = element_blank(), panel.grid.minor.x = element_blank())

### strip text
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    theme_bw() +
    facet_grid(vaccines ~ .) +
    theme(panel.grid.major.x = element_blank(), panel.grid.minor.x = element_blank(), strip.background = element_rect(fill = "gainsboro", color = NA), strip.text = element_text(size = 7))

### axis labels
ggplot(vac_subset) +
    geom_point(mapping = aes(x = date, y = total_vaccinations)) +
    theme_bw() +
    facet_grid(vaccines ~ .) +
    theme(panel.grid.major.x = element_blank(), panel.grid.minor.x = element_blank(), strip.background = element_rect(fill = "gainsboro", color = NA), strip.text = element_text(size = 7)) +
    xlab("Date") +
    ylab("Total vaccinations")

```

### Finalize example visualization

![*COVID19 vaccination rates over time*](https://www.dropbox.com/s/7bxebcgwgpdxax0/plot.png?dl=1){width=80%}

```{r example6, exercise=TRUE}
### add country label
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last) + 
    geom_text_repel(aes(x = date, y = total_vaccinations, label = country), size = 2, color = "black", direction = "y", data = vac_subset_last) +
    facet_grid(vaccines ~ .) + 
    scale_y_continuous(labels = label_comma()) +
    scale_colour_viridis(labels = label_comma()) +
    scale_fill_viridis(labels = label_comma()) +
    scale_shape_manual(values = shapes)
```

```{r example6-solution}
ggplot(vac_subset) +
    geom_path(mapping = aes(x = date, y = total_vaccinations, colour = daily_vaccinations_raw, group = country), size = 1) + 
    geom_point(mapping = aes(x = date, y = total_vaccinations, shape = country), size = 2, data = vac_subset_last, show.legend = FALSE) + 
    geom_text_repel(aes(x = date, y = total_vaccinations, label = country), size = 2, color = "black", direction = "y", data = vac_subset_last) +
    facet_grid(vaccines ~ .) + 
    scale_y_continuous(labels = label_comma()) +
    scale_colour_viridis(labels = label_comma()) +
    scale_fill_viridis(labels = label_comma()) +
    scale_shape_manual(values = shapes) + 
    ylab("Total vaccinations") +
    theme_bw() +
    theme(panel.grid.major.x = element_blank(), panel.grid.minor.x = element_blank(), strip.background = element_rect(fill = "gainsboro", color = NA), strip.text = element_text(size = 7))
```


## Exercises

### Exercise 1
Visualize the relationship between the rate of people vaccinated and the rate of people fully vaccinated. The visualization should include

* Points and lines as geoms 
* Aesthetic mapping to differentiate countries
* A dashed line to highlight x=y
* Fixed ratio coordinate system with equal units

```{r ex1, exercise=TRUE}
vac_subset
```
```{r ex1-hint-1}
ggplot(vac_subset, mapping = aes(x = people_vaccinated_per_hundred, y = people_fully_vaccinated_per_hundred)) +
    geom_point(aes(color = country)) +
    geom_line(aes(color = country)) 
```
```{r ex1-hint-2}
ggplot(vac_subset, mapping = aes(x = people_vaccinated_per_hundred, y = people_fully_vaccinated_per_hundred)) +
    geom_point(aes(color = country)) +
    geom_line(aes(color = country)) +
    geom_abline(linetype = "dashed")
```
```{r ex1-hint-3}
ggplot(vac_subset, mapping = aes(x = people_vaccinated_per_hundred, y = people_fully_vaccinated_per_hundred)) +
    geom_point(aes(color = country)) +
    geom_line(aes(color = country)) +
    geom_abline(linetype = "dashed") + 
    coord_fixed()
```
```{r ex1-solution}
ggplot(vac_subset, mapping = aes(x = people_vaccinated_per_hundred, y = people_fully_vaccinated_per_hundred)) +
    geom_point(aes(color = country)) +
    geom_line(aes(color = country)) +
    geom_abline(linetype = "dashed") + 
    coord_fixed()
```


```{r ex1-q1, echo=FALSE}
question("Which country has the highest vaccination rate?",
  answer("Israel"),
  answer("Belgium"),
  answer("Denmark", correct = TRUE),
  answer("Norway")
)
```

```{r ex1-q2, echo=FALSE}
question("Which of the following statements is true ",
  answer("Israel's strategy prioritized first shots over fully vaccinating at the start of the campaign", correct = TRUE),
  answer("Israel's strategy prioritized fully vaccinating over first shots over at the start of the campaign"),
  answer("Norway has not yet fully vaccinated all persons that started the vaccination", correct = TRUE)
)
```


### Exercise 2
Recreate this visualisation, and note down an observation you learned from it. 

![*Exercise 2*](https://www.dropbox.com/s/grf4aepmrjdj2lg/exercise_plot_1.png?dl=1){width=80%}

```{r ex2, exercise=TRUE}
vac_subset
```
```{r ex2-hint-1}
# Use the geom `geom_area` to create the stacked ribbons
```
```{r ex2-hint-2}
ggplot(vac_subset, aes(x = date, y = people_fully_vaccinated)) +
    geom_hline(yintercept = 0) +
    geom_area(aes(fill = country), colour = "white", size = 0.25)
```
```{r ex2-hint-3}
ggplot(vac_subset, aes(x = date, y = people_fully_vaccinated)) +
    geom_hline(yintercept = 0) +
    geom_area(aes(fill = country), colour = "white", size = 0.25) +
    scale_fill_brewer(name = "Country", palette = "Set1")
```
```{r ex2-solution}
ggplot(vac_subset, aes(x = date, y = people_fully_vaccinated)) +
    geom_hline(yintercept = 0) +
    geom_area(aes(fill = country), colour = "white", size = 0.25) +
    scale_fill_brewer(name = "Country", palette = "Set1") +
    scale_y_continuous(label = label_comma()) +
    xlab("Date") +
    ylab("People fully vaccinated") +
    theme(panel.grid.minor.x = element_blank(), panel.grid.major.x = element_blank())
```

### Exercise 3 
Recreate this visualisation, and note down an observation you learned from it. 

![*Exercise 3*](https://www.dropbox.com/s/xg6p4uq1318ty51/exercise_plot_2.png?dl=1){width=80%}

```{r ex3, exercise=TRUE, fig.height = 3.5, fig.width = 5}
vac_subset
```
```{r ex3-hint-1}
# Use the geom `geom_violin` to create the violin plots
```
```{r ex3-hint-2}
ggplot(vac_subset, mapping = aes(y = country, x = daily_vaccinations_per_million)) +
    geom_violin(scale = "width", fill = "gainsboro", color = NA)
```
```{r ex3-hint-3}
ggplot(vac_subset, mapping = aes(y = country, x = daily_vaccinations_per_million)) +
    geom_violin(scale = "width", fill = "gainsboro", color = NA) +
    geom_point(aes(fill = date), colour = "white", size = 1.5, position = position_jitter(height = 0.25), shape = 21, alpha = 0.9)
```
```{r ex3-solution}
ggplot(vac_subset, mapping = aes(y = country, x = daily_vaccinations_per_million)) +
    geom_violin(scale = "width", fill = "gainsboro", color = NA) +
    geom_point(aes(fill = date), colour = "white", size = 1.5, position = position_jitter(height = 0.25), shape = 21, alpha = 0.9) +
    scale_x_continuous(label = label_comma()) +
    scale_fill_viridis(name = "Date", option = "B", trans = "date") +
    xlab("Daily vaccinations per million") +
    ylab("Country") +
    theme_classic() +
    theme(axis.line.y = element_blank(), axis.ticks.y = element_blank())
```