---
title: "STAT 331 Week 4 Day 1 Handout"
format: html
embed-resources: true
---

```{r}
#| label: setup
#| message: false
#| echo: false

library(tidyverse)
library(liver)
```

## Data

::: callout-warning
For this handout, you need to download `imdb_data.Rdata` from the Week 4 Schedule Canvas page and save them on your computer.
:::

We will use 7 datasets that describe movies from [IMDb](https://www.imdb.com/). 

```{r}
#| echo: false
#| eval: true
#| fig-cap: "Relationship between data sets in IMDb movie data."
#| out-width: 90%
knitr::include_graphics("https://github.com/manncz/stat-331-s25/blob/main/slides/week-4/images/imdb_relational.png?raw=true")
```

Save the `imdb_data.Rdata` file to a reasonable directory, and change the file path below as needed to read it in.

```{r}
load(file = "data/imdb_data.Rdata")
```

## Mutating Joins

### `inner_join()`

```{r}
inner_join(directors_genres, 
           movies_directors, 
           by = "director_id")
```

a. When using one variable name for a join, does the variable name need to be in quotation marks (as a string)?


b. How does the number of observations in the result above compare to the `directors_genres` data and `movies_directors` data? Why does this make sense?



```{r}
inner_join(directors_genres, 
           directors, 
           by = join_by(director_id == id))
```

c. Why did we use `join_by()` in the join above? Note that the variable names do not need to be in quotation marks when using `join_by()`!




d. How does the number of observations in the result above compare to the `directors_genres` data and `directors` data? Why does this make sense?

### `left_join()`

```{r}
directors_genres |> 
  left_join(directors, 
            by = join_by(director_id == id))
```

e. Which directors will remain after the above join?


### `right_join()`

```{r}
directors_genres |> 
  right_join(directors, 
            by = join_by(director_id == id))
```

f. Which directors will remain after the above join?

### `full_join()`

```{r}
directors_genres |> 
  full_join(directors, 
            by = join_by(director_id == id))
```

f. Which directors will remain after the above join?


### Comparing joins

e. Why / when would you use each kind of join?

  - `inner_join()`
  - `left_join()`
  - `right_join()`
  - `full_join()`
  
f. What might be the cause of `NA` values when joining two datasets together?

## Filtering Joins

a. How are filtering joins different from mutating joins?



b. What is the equivalent to the below join, using the `filter()` verb instead?

```{r}
#| eval: false

actors |> 
  semi_join(roles,
            by = join_by(id == actor_id))
```


## Tidy Data

```{r}
data(cereal)
```

a. The `cereal` data is considered to be "wide" because...


b. How would we plot the mean cereal nutrients by shelf (as shown on the slides) with original cereal data?

```{r}


```


## Pivot

### `pivot_longer()`

a. What do you input in the `names_to` argument?


b. What do you input in the `values_to` argument?


a. Pivot the cereal data into long format in terms of the nutrient information. 

```{r}

```

### `pivot_wider()`

```{r}
mean_protein <- cereal |> 
  group_by(manuf, shelf) |> 
  summarize(mean_protein = mean(protein))
```

```{r}
mean_protein |> 
  arrange(shelf) |> 
  pivot_wider(id_cols = manuf,
              names_from = shelf,
              values_from = mean_protein)
```

a. Why are the names of the columns in the output above `1`, `2`, and `3`?

b. Why are there missing values in some of the columns?


c. How can we make the column names nicer? Specifically, I would like to make the column names something like "shelf_1".