R Code for Stats

.inf asked me to post this and I forgot about it until now. But as many of you will remember from the previous leagues, I've created a program that will compile a player's statistics over a given day range. Some of you have asked me for the code, so I figured I'd go ahead and share it publicly. First, if you don't have R downloaded on your computer, go ahead and do so now.

Link: www.r-project.org/

I'd also recommend downloading RStudio, which is an Integrated Development Environment (IDE) for R that makes it easier to run and understand.

Link: www.rstudio.com/

Next, run this code in R to create the functions that compile the statistics:

rm(list=ls())
options(stringsAsFactors = FALSE)
options(digits=3)
library('rvest')

path <- "http://tsfbl.atwebpages.com/"
# ADD archive/YEAR/ FOR PAST SEASONS

teams <- c("Celtics", "Heat", "Nets", "Knicks", "Magic", "76ers", "Wizards",
          "Hawks", "Hornets", "Bulls", "Cavaliers", "Pistons", "Pacers",
          "Bucks", "Raptors", "Mavericks", "Nuggets", "Rockets",
          "Timberwolves", "Spurs", "Jazz", "Grizzlies", "Warriors",
          "Clippers", "Lakers", "Suns", "Trail Blazers", "Kings",
          "SuperSonics")

colMax <- function(data) sapply(data, max, na.rm = TRUE)

get_player_stats <- function(player, team, start=1, end=200, per36=F){
 
 stats <- c()
 
 for(i in 1:length(team)){
   index <- which(team[i] == teams)
   url <- paste(path, "/rosters/roster", index, "sched.htm", sep="")
   webpage <- read_html(url)
   links <- html_attr(html_nodes(webpage, "a"), "href")
   links <- links[10:(length(links)-1)]
   
   for(j in 1:length(links)){
     url <- paste(path, substr(links[j], 3, nchar(links[j])), sep="")
     day <- as.numeric(substr(url, gregexpr("boxes/", url)[[1]][1]+6, gregexpr("-", url)[[1]][1]-1))
     
     if(day >= start & day <= end){
       full_box <- html_table(read_html(url), header=TRUE, fill=TRUE)
       
       if(colnames(full_box[[2]])[1] == team[i]){
         my_team_index <- 2
         opp_team_index <- 3
       } else{
         my_team_index <- 3
         opp_team_index <- 2
       }
       
       box <- full_box[[my_team_index]]
       box_team <- box[1:(nrow(box)-2), 1:(ncol(box)-1)]
       box_team[,3:16] <- lapply(box_team[,3:16], as.numeric)
       colnames(box_team)[6:7] <- c("X3P", "X3PA")
       if(colnames(box_team)[1] == team[i]){
         index <- which(box_team[,1] == player)
         if(length(index) > 0){
           colnames(box_team)[1] <- "Team"
           stats <- rbind(stats, box_team[index,])
         }
       }
     }
   }
 }
 
 if(is.null(stats)){
   cat("No stats found.\n")
 } else if(nrow(stats) == 0){
   cat("No stats found.\n")
 } else{
   
   stats <- cbind.data.frame(GM=nrow(stats), stats[, 3:16])
   
   cat("AVERAGES\n")
   averages <- colMeans(stats)

   print(round(averages, digits=1))
   
   if(per36 == TRUE){
     cat("\nPER 36\n")
     constant <- (36 / averages[2])
     for(i in 2:length(averages)){
       averages[i] <- constant * averages[i]
     }
     print(round(averages, digits=1))
   }
   
   cat("\nPERCENTAGES\n")
   print(c(sum(stats$FG) / sum(stats$FGA), sum(stats$X3P) / sum(stats$X3PA),
           sum(stats$FT) / sum(stats$FTA)))
 }
}

get_team_stats <- function(team, start = 1, end = 200){
 for(i in 1:length(team)){
   cat(team[i])
   cat("\n")
   get_player_stats(team[i], deparse(substitute(team)), start = start, end = end)
   cat("\n")
 }
}

If you're getting an error that R can't find the rvest package, type the following line of code first, and then re-run (you'll only have to do this once):

install.packages("rvest")

Once you've done that, you should be ready to go to compile the statistics all on your own. To run the selection for a given player, run the following code, for example:

get_player_stats("Kobe Bryant", "SuperSonics")

If you want to narrow down the date ranges, you can add in start and end arguments into the parameters. For example, if you wanted the same stats but only for Days 61-90:

get_player_stats("Kobe Bryant", "SuperSonics", start=61, end=90)

Finally, you can put in an argument to have the code print PER36 stats as well. For example:

get_player_stats("Kobe Bryant", "SuperSonics", per36=TRUE)

You can combine multiple arguments in one call if you wish.

As some of you may have seen, I've made it so the code can quickly do multiple people on the same team at once. In order to do that, you must first create a vector named the team name exactly. Then, you run the get_team_stats function. For example:

SuperSonics <- c("Kenny Anderson", "Kobe Bryant")
get_team_stats(SuperSonics)

Unfortunately, the naming conventions don't allow for the get_team_stats function to be run properly for the Trail Blazers (two word name) or the 76ers (name starts with a number). Please note that if you want to change the arguments in the get_team_stats function like above, you must edit the arguments as they're called in the get_player_stats function within get_team_stats.

Feel free to reach out to me with any questions about this if you run into trouble.