lecture_lab02.qmd

---
title: "Lecture Lab 2"
author: "Leon Eyrich Jessen"
format:
  revealjs:
    embed-resources: true
    theme: moon
    slide-number: c/t
    width: 1600
    height: 900
    mainfont: avenir
    logo: images/r4bds_logo_small.png
    footer: "R for Bio Data Science"
---

# Data Visualisation I

```{r}
library("ggplot2")
library("tibble")
library("dplyr")
```


<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## First: Make sure you're on track!

### Lab I Learning Objectives

-   Navigate the RStudio IDE and master the very basics of R
-   Create, edit and run a basic Quarto document
-   Explain why reproducible data analysis is important
-   Describe the components of a reproducible data analysis

### Where should you be by now?

-   In a group
-   On Slack
-   On the RStudio Cloud Server
-   On GitHub

*If any of this is not the case, see the `Getting Started` section on the course site*

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## What is Data Visualisation?

:::: {.columns}

::: {.column width="50%"}

-   Think about Hadley's and Hans' lectures, the book chapter and paper you read for today
-   Use the next 5 minutes to think and talk to the person next to you
-   Then, write your thoughts on Slack in the channel for Teaching lab 2

:::

::: {.column width="50%"}

![](images/what_is_data_viz.png)

:::
::::

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## On Data Visualisation

-   Depending on where your read, something along the lines of:
    -   *The graphical representation of information and data*
    -   *Uses visual elements like charts, graphs and maps*
    -   *Provide an accessible way to see and understand trends, outliers and patterns*
-   Data visualisation is your means to summarise and communicate key messages
    -   Within research
    -   Within industry
-   It is not easy!
    -   Anyone can make a plot, but an impactful data visualisation requires true skills

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## What is wrong here?

```{r}
#| echo: false
#| fig-align: center
tibble(performance = c(0.82, 0.84),
       method = c("Someone else", "me")) |>
  ggplot(aes(x = method, y = performance)) +
  geom_col() +
  coord_cartesian(ylim = c(0.80, 0.85))
```

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## Should have been

```{r}
#| echo: false
#| fig-align: center
tibble(performance = c(0.82, 0.84),
       method = c("Someone else", "me"))  |> 
  ggplot(aes(x = method, y = performance)) +
  geom_col() +
  geom_hline(yintercept = 0) +
  coord_cartesian(ylim = c(0, 1))
```

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## What's wrong here?

```{r}
#| echo: false
#| fig-align: center
set.seed(541880)
d = tibble(value = c(rnorm(20, 10, 2),
                 rnorm(20, 8, 2)),
       time = c(rep("t1", 20), rep("t2", 20)))
d |>
  group_by(time) |>
  summarise(mu = mean(value), sigma = sd(value)) |>
  ggplot(aes(x = time, y = mu)) +
  geom_col() +
  geom_errorbar(aes(ymin = mu - sigma,
                    ymax = mu + sigma), width = 0.1)
```

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## Should have been

```{r}
#| echo: false
#| fig-align: center
d |> 
  ggplot(aes(x = time, y = value)) +
  geom_boxplot()
```

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## Or maybe even

```{r}
#| echo: false
#| fig-align: center
d |> 
  ggplot(aes(x = time, y = value)) +
  geom_violin(scale = "width") +
  geom_boxplot(width = 0.2)
```

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## What is this?

![](images/bad_viz_01.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->

## Or this?

![](images/bad_viz_02.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->

## Or this?

![](images/bad_viz_03.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->

## Or this?

![](images/bad_viz_04.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->

## Or this?

![](images/bad_viz_05.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->

## Or this?

![](images/bad_viz_06.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->

## Or this?

![](images/bad_viz_07.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## ...and then there is this one?!?

![](images/bad_viz_17.jpeg){fig-align="center" width="80%"}


<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## ...and then there is this one?!?

![](images/bad_viz_17_mod.jpeg){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## Or this?

![](images/bad_viz_08.png){fig-align="center" width="80%"}

-   Don't even know where to begin with this one...

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## In Summary on Data Visualisation

-   Think carefully about *exactly* what it is, you want to communicate with your visualisation

-   Remove redundant information

-   Be honest and show your data

-   Less is more, do not cram 3 plots into 1

-   Do not make a fancy, but information deprived plot

-   Think about colour choice - Separation, but also quite a few people are colour blind

-   Today is meant as in intro, data visualisation will be an integrated part of the rest of the course

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## ggplot, the Grammar of Graphics

### The basic syntax of ggplot

-   In a new code chunk in your Quarto document, input and run:

```{{r}}
ggplot(data = my_data,
       mapping = aes(x = v1, y = v2)) +
  geom_point()
```

-   Define your data

-   Map variables in your plot to your visualisation

-   Choose a graphical representation

-   Let us look at that in another way

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## ggplot, the Grammar of Graphics

![](images/vis_ggplot_components.png){fig-align="center" width="80%"}

<!--# ---------------------------------------------------------------------- -->
<!--# SLIDE ---------------------------------------------------------------- -->
<!--# ---------------------------------------------------------------------- -->
## Example - Scatter plot

```{r}
#| echo: true
#| fig-align: center
#| fig-height: 3
library("tidyverse")
ggplot(data = datasets::Puromycin,
       mapping = aes(x = conc,
                     y = rate)) +
  geom_point()
```

# Break and the it is time for exercises...