s1-exercises-with-answers.Rmd

---
title: "Session 1 Exercises (with answers)"
author: "Ben Raymond, Adrien Ickowicz"
output:
  html_document:
    toc: false
    theme: flatly
    highlight: zenburn
    css: "extra/exercises.css"
---

```{r chunkopts, eval = TRUE, echo = FALSE}
knitr::opts_chunk$set(message = FALSE, warning = FALSE)
options(width=80)
options(tibble.width = 80, tibble.print_max = 10, tibble.print_min = 10)

```

# Preparation

Load some packages:

```{r setup}
library(tidyverse)
library(lubridate)

```

And read in our data (set scores from the 2019 D1 Women's NCAA competition, scraped from the NCAA web site).

```{r data}
x <- read_csv("example_data/NCAA_D1W_2019.csv")

glimpse(x)

```

# Exercises

### 1. What is the date range of the matches?

<details>
  <summary>Click for solution</summary>

```{r}
range(x$date)
```

</details>


### 2. How many matches were played?

<details>
  <summary>Click for solution</summary>

```{r}
length(unique(x$match_id))
```

</details>

### 3. How many teams?

<details>
  <summary>Click for solution</summary>

```{r}
length(unique(c(x$home_team, x$visiting_team)))
```

</details>

### 4. Count the number of matches that went to 3, 4, and 5 sets

<details>
  <summary>Click for solution</summary>

```{r}
x %>% group_by(match_id) %>% slice(1) %>% ungroup %>% ## one row per match
    count(match_n_sets) ## count the number of times each value of match_n_sets appears

## or
x %>% distinct(match_id, match_n_sets) %>%
    count(match_n_sets)
```

</details>

### 5. On average, how many sets were played per match?

<details>
  <summary>Click for solution</summary>

```{r}
x %>% group_by(match_id) %>% slice(1) %>% ungroup %>%
    dplyr::summarize(mean_n_sets = mean(match_n_sets))

## or
nrow(x) / length(unique(x$match_id))
```

</details>

### 6. How many matches did Weber State play?

<details>
  <summary>Click for solution</summary>

```{r}
x %>% filter(home_team == "Weber State" | visiting_team == "Weber State") %>%
    summarize(n_matches = length(unique(match_id)))

```
or perhaps

```{r}
x %>% filter(home_team == "Weber State" | visiting_team == "Weber State") %>%
    distinct(match_id) %>% nrow

```

or even

```{r}
length(unique(x$match_id[x$home_team == "Weber State" | x$visiting_team == "Weber State"]))

```

</details>

### 7. List the top 10 teams in terms of the most matches played

<details>
  <summary>Click for solution</summary>

```{r}
bind_rows(x %>% distinct(match_id, team = home_team),
          x %>% distinct(match_id, team = visiting_team)) %>%
    count(team) %>% ## count the number of times each team name appears
    dplyr::arrange(desc(n)) %>% ## arrange in decreasing count order
    head(10) ## first 10 rows

```

</details>

### 8. Can you re-create the `margin` column, using the other columns in the data set?

<details>
  <summary>Click for solution</summary>

```{r}
## add some extra columns
x <- x %>% mutate(margin = abs(home_team_score - visiting_team_score))

```

</details>


### 9. What was the average winning margin of a set?

<details>
  <summary>Click for solution</summary>

```{r}
mean(x$margin)
```

</details>


### 10. Does that vary by set number 1--5?

<details>
  <summary>Click for solution</summary>

```{r}
x %>% group_by(set_number) %>% dplyr::summarize(mean_winning_margin = mean(margin))
```

</details>

### 11. Plot a histogram of the score margin (all sets combined)

<details>
  <summary>Click for solution</summary>

```{r}
ggplot(x, aes(x = margin)) + geom_histogram(binwidth = 1) +
    labs(x = "Score margin", y = "Number of matches")
```

</details>


### 12. Same thing with each set shown separately

<details>
  <summary>Click for solution</summary>

```{r}
ggplot(x, aes(x = margin)) + geom_histogram(binwidth = 1) +
    facet_wrap(~set_number) +
    labs(x = "Score margin", y = "Number of matches")

## or
ggplot(x, aes(x = margin, group = set_number, fill = as.factor(set_number))) +
    geom_histogram(binwidth = 1, position = "dodge") + scale_fill_discrete(name = "Set number") +
    labs(x = "Score margin", y = "Number of matches")
```

</details>

### 13. Find the most one-sided set result

<details>
  <summary>Click for solution</summary>

```{r}
x %>% slice_max(margin)

## or
x[which(x$margin == max(x$margin)), ]
```

</details>

### 14. Is there a home team advantage? Find the match win rate for the home team (ignore the fact that some of these matches were played at neutral venues --- just take the `home_team` and `visiting_team` column names on face value)

<details>
  <summary>Click for solution</summary>

```{r}
x <- x %>% group_by(match_id) %>% mutate(match_won_by = case_when(home_team_set_score > visiting_team_set_score ~ home_team,
                                                                  TRUE ~ visiting_team)) %>%
    ungroup

## but
x %>% dplyr::filter(match_won_by == visiting_team & visiting_team_set_score <= home_team_set_score)

## so
x <- x %>% group_by(match_id) %>% mutate(match_won_by = case_when(home_team_set_score > visiting_team_set_score ~ home_team,
                                                                  visiting_team_set_score > home_team_set_score ~ visiting_team)) %>%
    ungroup

x <- x %>% mutate(home_team_won_match = case_when(match_won_by == home_team ~ TRUE,
                                                  match_won_by == visiting_team ~ FALSE))

## and finally
x %>% group_by(match_id) %>% slice(1) %>% ungroup %>%
    dplyr::summarize(p_home_win = mean(home_team_won_match, na.rm = TRUE))

## or
x %>% distinct(match_id, home_team_won_match) %>%
    dplyr::summarize(p_home_win = mean(home_team_won_match, na.rm = TRUE))
```

</details>

<div id="footer"></div>