Skip to content

Instantly share code, notes, and snippets.

@isaactpetersen
Last active August 29, 2015 14:04
Show Gist options
  • Select an option

  • Save isaactpetersen/80586cfbbd7e90e807e9 to your computer and use it in GitHub Desktop.

Select an option

Save isaactpetersen/80586cfbbd7e90e807e9 to your computer and use it in GitHub Desktop.
#Libraries
library("XML")
#Specify info to scrape
years <- 2011:2013
weeks <- 17
#Scrape data
qb <- list()
rb1 <- list()
rb2 <- list()
rb3 <- list()
wr1 <- list()
wr2 <- list()
wr3 <- list()
wr4 <- list()
wr5 <- list()
wr6 <- list()
pb <- txtProgressBar(min = 1, max = weeks, style = 3)
for(i in 1:weeks){
setTxtProgressBar(pb, i)
qb[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&game_type=R&league_id=&team_id=&opp_id=&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_location=&game_result=&handedness=&is_active=&is_hof=&c1stat=pass_att&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=pass_att", sep=""), stringsAsFactors = FALSE)$stats
rb1[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&game_type=R&league_id=&team_id=&opp_id=&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_location=&game_result=&handedness=&is_active=&is_hof=&c1stat=rush_att&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rush_yds", sep=""), stringsAsFactors = FALSE)$stats
rb2[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&league_id=&team_id=&opp_id=&game_type=R&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_month=&game_location=&game_result=&is_active=&handedness=&is_hof=&c1stat=rush_att&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rush_yds&order_by_asc=&offset=100", sep=""), stringsAsFactors = FALSE)$stats
rb3[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&league_id=&team_id=&opp_id=&game_type=R&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_month=&game_location=&game_result=&is_active=&handedness=&is_hof=&c1stat=rush_att&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rush_yds&order_by_asc=&offset=200", sep=""), stringsAsFactors = FALSE)$stats
wr1[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&game_type=R&league_id=&team_id=&opp_id=&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_location=&game_result=&handedness=&is_active=&is_hof=&c1stat=rec&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rec_yds", sep=""), stringsAsFactors = FALSE)$stats
wr2[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&league_id=&team_id=&opp_id=&game_type=R&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_month=&game_location=&game_result=&is_active=&handedness=&is_hof=&c1stat=rec&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rec_yds&order_by_asc=&offset=100", sep=""), stringsAsFactors = FALSE)$stats
wr3[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&league_id=&team_id=&opp_id=&game_type=R&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_month=&game_location=&game_result=&is_active=&handedness=&is_hof=&c1stat=rec&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rec_yds&order_by_asc=&offset=200", sep=""), stringsAsFactors = FALSE)$stats
wr4[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&league_id=&team_id=&opp_id=&game_type=R&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_month=&game_location=&game_result=&is_active=&handedness=&is_hof=&c1stat=rec&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rec_yds&order_by_asc=&offset=300", sep=""), stringsAsFactors = FALSE)$stats
wr5[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&league_id=&team_id=&opp_id=&game_type=R&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_month=&game_location=&game_result=&is_active=&handedness=&is_hof=&c1stat=rec&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rec_yds&order_by_asc=&offset=400", sep=""), stringsAsFactors = FALSE)$stats
wr6[[i]] <- readHTMLTable(paste("http://www.pro-football-reference.com/play-index/pgl_finder.cgi?request=1&match=game&year_min=", head(years, 1), "&year_max=", tail(years, 1), "&season_start=1&season_end=-1&age_min=0&age_max=99&league_id=&team_id=&opp_id=&game_type=R&game_num_min=0&game_num_max=99&week_num_min=", i, "&week_num_max=", i, "&game_day_of_week=&game_month=&game_location=&game_result=&is_active=&handedness=&is_hof=&c1stat=rec&c1comp=gt&c1val=1&c2stat=&c2comp=gt&c2val=&c3stat=&c3comp=gt&c3val=&c4stat=&c4comp=gt&c4val=&order_by=rec_yds&order_by_asc=&offset=500", sep=""), stringsAsFactors = FALSE)$stats
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment