---
title: "EC306 — R and Quarto Practice File"
author: "Your name, your student number"
date: today
format: pdf
---

<!--
  HOW TO USE THIS FILE
  --------------------
  1. Download BOTH this file and the data file ec306lfs.csv from the
     course website, and put them in the SAME folder — for example a
     folder called EC306 on your Desktop. Keep the .qmd extension.
  2. Open this file in RStudio (File > Open File).
  3. Press the Render button at the top of the editor. RStudio will run
     every code chunk and produce a PDF next to this one.
  4. Once that works, read through and change things. Break it, fix it.
     That is how you learn this.

  IF RENDERING FAILS with an error mentioning LaTeX, tlmgr, or a
  missing .sty file, the PDF machinery is not installed yet. Run this
  once in the Console, then render again:

      install.packages("tinytex")
      tinytex::install_tinytex()
-->

## Before you start

This file is a working example of a Quarto document. It contains three
kinds of content, and it is worth learning to recognise them:

1.  The **YAML header** at the very top, between the `---` lines. It
    sets the title, the author, and the output format.
2.  **Text**, like this paragraph, written in Markdown.
3.  **Code chunks**, which start with ```` ```{r} ```` and end with
    ```` ``` ````. Everything inside is R code and gets run when you
    render.

To make a new chunk, press **Ctrl+Alt+I** (**Cmd+Option+I** on a Mac).

To run a single chunk without rendering the whole document, press the
small green arrow at its top right.

## Setup

Every document starts by loading the packages it needs. If this chunk
gives you an error saying there is no package called 'tidyverse', run
`install.packages("tidyverse")` once in the Console, then try again.

```{r}
#| label: setup
#| message: false
library(tidyverse)
```

## The data

Now read in the data. This is the same Labour Force Survey extract used
in the tutorial: 500 working-age people, with their age, sex, education,
labour force participation, and hourly wage.

`read_csv()` looks for the file in the same folder as this document, so
if you get an error saying the file cannot be opened, that is almost
certainly the problem.

```{r}
#| label: load-data
#| message: false
lfs <- read_csv("ec306lfs.csv")
```

## Look at the data first

Never analyse data you have not looked at.

```{r}
glimpse(lfs)
```

```{r}
head(lfs)
```

A continuous variable like `wage` is summarised with means and
percentiles:

```{r}
summary(lfs$wage)
```

A categorical variable like `educ` is summarised by counting:

```{r}
count(lfs, educ)
```

## Summarise by group

This is the workhorse of applied work: an average, computed separately
for each group.

```{r}
lfs |>
  group_by(educ) |>
  summarise(
    n         = n(),
    mean_wage = mean(wage)
  )
```

## Make a graph

Remember to label your axes. Unlabelled graphs lose marks.

```{r}
#| label: fig-wage-educ
#| fig-cap: "Hourly wages by education level"
#| fig-width: 6
#| fig-height: 3.5
ggplot(lfs, aes(x = educ, y = wage)) +
  geom_boxplot() +
  labs(x = NULL, y = "Hourly wage ($)")
```

## Run a regression

Labour economists usually model the log of wages, so that coefficients
read as approximate percentage effects.

```{r}
lfs <- lfs |>
  mutate(lnwage = log(wage))

model <- lm(lnwage ~ yrschool + age, data = lfs)

summary(model)
```

And present it as a table rather than as console output:

```{r}
#| message: false
library(modelsummary)

modelsummary(model, stars = TRUE)
```

## Your turn

Try each of these by adding a chunk underneath. If something breaks,
read the first line of the error message, then check spelling,
capitalisation, and brackets.

1.  Count how many people in the data are under 30. (Hint: `filter()`,
    then `nrow()`.)
2.  Compute the mean wage separately for men and women, using the
    `female` variable.
3.  Make a histogram of `wage` using `geom_histogram()`.
4.  Add `female` to the regression and look at what happens to the
    coefficients.

```{r}
# your code here

```

## When you are finished

Press Render one last time and check that it completes without errors.
A document that will not render is a document you cannot submit.