R markdown notebooks accompanying yt screencast

2021-06-19-nba-21-bbr.Rmd

---
title: "Basketball Reference"
layout: post
categories:
  - notebook
tags:
  - nba, sports
output: 
   html_document
#  md_document:
#    variant: markdown_github+backtick_code_blocks
#    preserve_yaml: true
#    toc: false
#    fig_retina: 2
  
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

# Compare '21 Brooklyn Nets team performance based on star players


```{r include=FALSE}
library(rvest)
library(cowplot)
library(tidyverse)
library(xml2)
```


```{r}
roster <- "https://www.basketball-reference.com/teams/BRK/2021.html"

html <- read_html(roster)
tables <- html %>% html_nodes("table") %>% html_table()
df <- tables[1] %>% data.frame()

players <- html %>% html_nodes("table") %>% xml_find_all(".//a") %>% xml_attrs()
gamelogs <- unlist(players)
pgl <- gamelogs[grep("gamelog", gamelogs)]  %>% unique() %>% list() 
pgl <- pgl[[1]]

playoffs_pl <- pgl[c(1,2,3,4,6,9,10,14,12,15,18)]
```

```{r }

get_player_url <- function(path) {
  paste("https://www.basketball-reference.com/",path,sep="")
}

player_urls <- unlist(lapply(FUN=get_player_url, playoffs_pl))

get_data <- function(player_url) {
  html <- read_html(player_url)
  tables <- html %>% html_nodes("table") %>% html_table()
  df <- tables[8] %>% data.frame()
  return(df)
}

```

```{r}

for (p in player_urls) {
   pname <- str_split(p, "/")[[1]][7]
   data <- get_data(p)
   assign(pname, data) 
}
```

```{r}
plist <- list()
for (p in player_urls) {
   pname <- str_split(p, "/")[[1]][7]
   plist <- c(plist, pname)   # Add element to list
}
unlist(plist)

```

```{r}

i = 1
for (p in plist) {
  
  if (i == 1) {
      data = get(p)
      data$player <- p
        
  } else {
      df = get(p)
      df$player <- p
      data = rbind(data, df) 
  }
  i = i + 1
}
  
```

```{r}
table(data$player)

data %>% filter(Rk != "Rk") %>% nrow()
nrow(data)

data <- data %>% filter(Rk != "Rk") 
```


```{r}
nets_games_url <- 'https://www.basketball-reference.com/teams/BRK/2021_games.html'
html <- read_html(nets_games_url)
tables <- html %>% html_nodes("table") %>% html_table()

nets_regular_season <- tables[[1]] %>% data.frame()
nets_playoffs <- tables[[2]] %>% data.frame()

games <- nets_regular_season %>% select(G, Date, Var.6, Opponent, Var.8, Tm, Opp)
names(games) <- c("G", "Date", "Home", "Opponent", "WinLoss", "Team", "Opp")
```

```{r}

games$date <- apply(games, 1, FUN = function(x) {substr(x[c("Date")], 6, 20) })
games$date <- as.Date(games$date, c("%B %d, %Y"))

glimpse(games)

games$Team <- as.integer(games$Team)
games$Opp <- as.integer(games$Opp)

games$diff <- games$Team - games$Opp
```

```{r warning=FALSE}
hist(games$diff)
games$home <- games$Home != '@'
table(games$home)
#qplot(games$date, games$diff)
games %>% ggplot(.) + geom_point(aes(y=diff, x=date)) + facet_grid(home ~ .) + geom_smooth(aes(x=date, y=diff))
```

```{r}
data$date <- as.Date(data$Date, format="%Y-%m-%d")
dfm <- games %>% left_join(data, by=c("date"="date"))

mp <- dfm %>% select(date, MP, player, WinLoss, diff)
mp$min <- as.difftime(mp$MP, format = "%M:%S", units = "mins")

mp %>% filter(!is.na(player)) %>% ggplot(.) + geom_point(aes(x=date, y=min, colour=player)) +
  facet_grid(WinLoss ~ .) + theme_cowplot()

mp %>% filter(!is.na(player)) %>% ggplot(.) + geom_point(aes(x=date, y=min, colour=player)) +
  facet_grid(player ~ .) + theme_cowplot() + labs(y="minutes played")

p1 <- mp %>% filter(!is.na(player) & WinLoss == "W") %>% ggplot(.) + geom_point(aes(x=date, y=min, colour=player)) + facet_grid(player ~ .) + theme_cowplot() + theme(legend.position = "none") + labs(y="minutes played")

p2 <- mp %>% filter(!is.na(player) & WinLoss == "L") %>% ggplot(.) + geom_point(aes(x=date, y=min, colour=player)) + facet_grid(player ~ .) + theme_cowplot()

plot_grid(p1, p2)


mp %>% filter(!is.na(player)) %>% ggplot(.) + geom_point(aes(x=date, y=min, colour=WinLoss)) + facet_grid(player ~ .) + theme_cowplot()


```