In systats/lolR: Data Sourcing

• DJT popularity likes twitter

Packages

pacman::p_load(dplyr, ggplot2, rvest, purrr, xml2, stringr)

Base Scraper

Helper Functions

base_url <- "https://lol.gamepedia.com"

str_extract_href <- function(x, append = F){
  x <- x %>% 
    as.character() %>%
    stringr::str_extract('href=\".*?\"') %>% 
    stringr::str_replace_all('href=|\"', '')

  if(append) x <- paste0(base_url, x)
  return(x)
}

Example Data

example_url <- paste0("https://lol.gamepedia.com/Special:RunQuery/MatchHistoryTournament?MHT%5Btournament%5D=Concept:", "MSI%202018", "%20Main%20Event&MHT%5Btext%5D=Yes&pfRunQueryFormName=MatchHistoryTournament")

# example_url <- "https://lol.gamepedia.com/2018_Mid-Season_Invitational/Main_Event/Match_History"
# 

example_html <- example_url %>% 
  xml2::read_html()

x <- example_html %>% 
  html_nodes(".wikitable") %>% 
  html_children() %>% 
  .[3:(length(.)-1)] %>% 
  .[25] # map index

Match Data

get_match_data(x)%>% glimpse

Player Stats

get_player_stats(x) %>% glimpse

Match Stats

Compare to Oracleelex.

load("data/match_stats.Rdata")
match_stats %>% glimpse()

stats_names <- example_html %>% 
  html_nodes(".wikitable") %>% 
  html_children() %>% 
  .[2] %>% 
  html_text() %>% 
  stringr::str_split("\\n+") %>% 
  unlist() %>% 
  stringr::str_trim()

stats_names

stats_names[12:30] %>% dput

get_match_stats(x) %>% glimpse()

TODO

• Runes incl. Position and Items (by Tournament) and TEam abbrevs
• Champion Aggregates
• Player Aggregates
• Player Profiles
  • Champion by and Tournament Plyer Stats
  • Match History
• Team Rosters (no clue)
• Picks and Bans

Meta Data

get_match_meta(x) %>% glimpse()

combine stats

get_match_history(x) %>% glimpse

Map Over Games

nn <- example_html %>% 
  html_nodes(".wikitable") %>% 
  html_children() %>% 
  .[3:(length(.)-1)] %>% 
  #.[1] # map index
  map(get_match_history_safely)


nn %>% 
  map("result") %>% 
  bind_rows()

Map Over Tournaments

tournament_url <- "https://lol.gamepedia.com/Special:RunQuery/MatchHistoryTournament?MHT%5Btournament%5D=Concept:CK%202017%20Spring%20Playoffs&MHT%5Btext%5D=Yes&pfRunQueryFormName=MatchHistoryTournament"

tournament_list <- read_html(tournament_url) %>%
  html_node(".inputSpan") %>%
  html_children() %>%
  html_children() %>%#
  html_text() %>%
  .[-1] %>%
  str_replace_all(., "\\s", "%20") %>%
  paste0(base_url,"/", "Special:RunQuery/MatchHistoryTournament?MHT%5Btournament%5D=", ., "&MHT%5Btext%5D=Yes&pfRunQueryFormName=MatchHistoryTournament")

get_tournament_matches_pro <- progressively(get_tournament_matches, 10)

gamepedia_list <- tournament_list[1:10] %>% 
  as.list() %>% 
  map(~{
    .x %>% 
      xml2::read_html() %>% 
      get_tournament_matches_pro
  })

#gamepedia_list
#gamepedia_list %>% glimpse(max.depth = 2)

gamepedia <- gamepedia_list %>% 
  map(~{
    .x %>% 
      map("result") %>% 
      bind_rows()
  }) %>% 
  bind_rows 

gamepedia %>% 
  count(is.na(game_hash))

table(is.na(gamepedia))

systats/lolR documentation built on May 31, 2019, 6:17 p.m.