WNBA Table Manipulation.Rmd

---
title: "WNBA Table Manipulation"
output:
  html_document: default
  pdf_document: default
---
```{r}

# Read in the data and create a base and ridge regressed model

points <- read.csv("C://Users/Brad Hymas/Desktop/School/Senior/Spring 2021/IS 590R/point_diff_2019_updated.csv", header = TRUE)
library(tidyverse)
y <- points$point_diff
x <- points %>% dplyr::select(-point_diff, -game_id, -home_possession) %>% as.matrix()
library(MASS)
ridge.model <- lm.ridge(y~-1+x, lambda = 2000)
model <- lm(y~-1+x)
ridge.coefs <- ridge.model$coef * 100
coefs <- model$coef * 100

# Variance of the coefficients

lm_var <- diag(vcov(model))* 1000
s2 <- 1/(nrow(x) - ncol(x)) * t(y - x %*% matrix(ridge.model$coef)) %*% (y - x %*% matrix(ridge.model$coef))
ridge_var <- solve(t(x) %*% x + 2000* diag(ncol(x))) %*% t(x) %*% x %*% solve(t(x) %*% x + 2000 * diag(ncol(x)))
finalRidgeVar <- as.vector(s2) * diag(ridge_var) * 1000
```

```{r}

# Determining player interactions

test <- as.matrix(points)
transTest<- t(test)
int <- transTest %*% test
finalInt <- int[3:(nrow(int) - 1), 3:(ncol(int) - 1)]

# Create a list of the amount of possessions played by each player

title <- colnames(finalInt)
possessions <- finalInt[1,1]

for (i in 2:nrow(finalInt)) {
  possessions <- c(possessions, finalInt[i, i])
}

# Create a data frame for the players and the number of possessions played, sorted decreasing

temp <- data.frame("Player Name" = title, "Possessions Played" = possessions)
index <- order(temp[,2], decreasing = TRUE)
playerPossessions <- temp[index,]

```


```{r}
# Per game stats

player2019 <- read.csv("C://Users/Brad Hymas/Desktop/School/Senior/Spring 2021/IS 590R/Players_2019.csv")
player2019test <- player2019[,c(2, 4:6, 8:length(player2019))]

# Defensive rebounds per player per game
drb <- player2019test[,18] - player2019test[,17]

# Minutes played per game
mppg <- player2019test[,8] / player2019test[,6]

# Field goals made per game
fgpg <- player2019test[,9] / player2019test[,6]

# Field goals attempted per game
fgapg <- player2019test[,10] / player2019test[,6]

# Two-pointers made per game
twoppg <- player2019test[,11] / player2019test[,6]

# Two-pointers attempted per game
twopapg <- player2019test[,12] / player2019test[,6]

# Three-pointers made per game
threeppg <- player2019test[,13] / player2019test[,6]

# Three-pointers attempted per game
threepapg <- player2019test[,14] / player2019test[,6]

# Free throws made per game
ftpg <- player2019test[,15] / player2019test[,6]

# Free throws attempted per game
ftapg <- player2019test[,16] / player2019test[,6]

# Offensive rebounds per game
orbpg <- player2019test[,17] / player2019test[,6]

# Defensive rebounds per game
drbpg <- drb / player2019test[,6]

# Total rebounds per game
trbpg <- player2019test[,18] / player2019test[,6]

# Assists per game
apg <- player2019test[,19] / player2019test[,6]

# Steals per game
spg <- player2019test[,20] / player2019test[,6]

# Blocks per game
bpg <- player2019test[,21] / player2019test[,6]

# Turnovers per game
tpg <- player2019test[,22] / player2019test[,6]

# Personal fouls per game
pfpg <- player2019test[,23] / player2019test[,6]

# Points per game
ppg <- player2019test[,24] / player2019test[,6]

# Possessions per game
pog <- (possessions / player2019test[,6])

```

```{r}

# Per 36 minutes stats

# Defensive rebounds per 36 minutes
drbm <- (drb / player2019test[,8]) * 36

# Field goals made per 36 minutes
fgm <- (player2019test[,9] / player2019test[,8]) * 36

# Field goals attempted per 36 minutes
fgam <- (player2019test[,10] / player2019test[,8]) * 36

# Two-pointers made per 36 minutes
twom <- (player2019test[,11] / player2019test[,8]) * 36

# Two-pointers attempted per 36 minutes
twoam <- (player2019test[,12] / player2019test[,8]) * 36

# Three-pointers made per 36 minutes
threem <- (player2019test[,13] / player2019test[,8]) * 36

# Three-pointers attempted per 36 minutes
threeam <- (player2019test[,14] / player2019test[,8]) * 36

# Free throws made per 36 minutes
ftm <- (player2019test[,15] / player2019test[,8]) * 36

# Free throws attempted per 36 minutes
ftam <- (player2019test[,16] / player2019test[,8]) * 36

# Offensive rebounds per 36 minutes
orbm <- (player2019test[,17] / player2019test[,8]) * 36

# Total rebounds per 36 minutes
trbm <- (player2019test[,18] / player2019test[,8]) * 36

# Assists per 36 minutes
am <- (player2019test[,19] / player2019test[,8]) * 36

# Steals per 36 minutes
sm <- (player2019test[,20] / player2019test[,8]) * 36

# Blocks per 36 minutes
bm <- (player2019test[,21] / player2019test[,8]) * 36

# Turnovers per 36 minutes
tm <- (player2019test[,22] / player2019test[,8]) * 36

# Personal fouls per 36 minutes
pfm <- (player2019test[,23] / player2019test[,8]) * 36

# Points per 36 minutes
pm <- (player2019test[,24] / player2019test[,8]) * 36

# Possessions per 36 minutes
pom <- (possessions / player2019test[,8]) * 36

```

```{r}

# Adding per game data to columns 

player2019test['APG'] <- apg
player2019test['BPG'] <- bpg
player2019test['DRB'] <- drb
player2019test['DRBPG'] <- drbpg
player2019test['FGPG'] <- fgpg
player2019test['FGAPG'] <- fgapg
player2019test['FTPG'] <- ftpg
player2019test['FTAPG'] <- ftapg
player2019test['MPG'] <- mppg
player2019test['ORBPG'] <- orbpg
player2019test['PFPG'] <- pfpg
player2019test['PPG'] <- ppg
player2019test['SPG'] <- spg
player2019test['X3PPG'] <- threeppg
player2019test['X3PAPG'] <- threepapg
player2019test['TPG'] <- tpg
player2019test['TRBPG'] <- trbpg
player2019test['X2PPG'] <- twoppg
player2019test['X2PAPG'] <- twopapg

# Adding per 36 minutes data to columns

player2019test['FGPM'] <- fgm
player2019test['FGAPM'] <- fgam
player2019test['X2PPM'] <- twom
player2019test['X2PAPM'] <- twoam
player2019test['X3PPM'] <- threem
player2019test['X3PAPM'] <- threeam
player2019test['FTPM'] <- ftm
player2019test['FTAPM'] <- ftam
player2019test['ORBPM'] <- orbm
player2019test['TRBPM'] <- trbm
player2019test['DRBPM'] <- drbm
player2019test['ASPM'] <- am
player2019test['SPM'] <- sm
player2019test['BPM'] <- bm
player2019test['TPM'] <- tm
player2019test['PFPM'] <- pfm
player2019test['PPM'] <- pm

# Adding player possession data

player2019test['POPG'] <- pog
player2019test['POS'] <- possessions
player2019test['POPM'] <- pom

# Reorganize the columns so the per game columns are next to the corresponding season totals

player2019pg <- player2019test[, c(1:8, 38, 9, 34, 10, 35, 11, 47, 12, 48, 13, 43, 14, 44, 15, 36, 16, 37, 32, 33, 17, 39, 18, 46, 19, 30, 20, 42, 21, 31, 22, 45, 23, 40, 24, 41, 67, 66, 25:29)]

# Reorganize the columns so the per 36 minutes columns are next to the corresponding season totals

player2019pm <- player2019test[, c(1:9, 49, 10, 50, 11, 51, 12, 52, 13, 53, 14, 54, 15, 55, 16, 56, 17, 57, 18, 58, 32, 59, 19, 60, 20, 61, 21, 62, 22, 63, 23, 64, 24, 65, 67, 68, 25:29)]
```

```{r}

# Add the coefficients from the base model and ridge regressed model to the table for both tables

player_2019_info_per_game <- player2019pg %>% arrange(Player) %>% mutate(APM = coefs, RAPM = ridge.coefs)
salaries <- read.csv("C://Users/Brad Hymas/Desktop/School/Senior/Spring 2021/IS 590R/WNBA player salary 20192.csv")
players_2019_per_game <- player_2019_info_per_game %>% left_join(salaries %>% dplyr::select(Player, Salary), by = "Player")
write.csv(players_2019_per_game, 'shiny_data.csv')

player_2019_info_per_minutes <- player2019pm %>% arrange(Player) %>% mutate(APM = coefs, RAPM = ridge.coefs)
players_2019_per_minutes <- player_2019_info_per_minutes %>% left_join(salaries %>% dplyr::select(Player, Salary), by = "Player")
write.csv(players_2019_per_minutes, 'shiny_data.csv')

```