userPicks.R

#' Detailed data for a user's player picks for a gameweek in the current FPL season
#'
#' Returns a tibble containing player data for a given user's selection for a gameweek during the current FPL season (season has to be in-play).

library(dplyr)

userPicks2 <- function(user_id, gameweek) {
  
  # make the input numeric
  user_id <- as.numeric(user_id)
  
  # check the input is numeric, stop if not
  if (!is.numeric(gameweek))
    stop("gameweek isn't numeric format.")
  
  # check the input is in range, stop if not
  if (!gameweek %in% 1:38)
    stop("gameweek isn't in expected range (1-38).")
  
  # get player/teams list
  players <- jsonlite::read_json("https://fantasy.premierleague.com/drf/bootstrap-static", simplifyVector = TRUE)
  teams <- subset(players$teams, select = c(id, name, strength, code))
  players <- subset(players$elements, select = c(id, web_name, team_code))
  
  # get user picks data
  data <- jsonlite::read_json(paste0("https://fantasy.premierleague.com/drf/entry/", user_id, "/event/", gameweek,
                                     "/picks"), simplifyVector = TRUE)
  picks <- data$picks
  
  # get player points data
  points <- list()
  
  for (i in picks$element) {
    data <- jsonlite::read_json(paste0("https://fantasy.premierleague.com/drf/element-summary/", i), simplifyVector = TRUE)
    
    data.2 <- subset(data$history, round == gameweek, select = c(id, kickoff_time, team_h_score:value, selected:transfers_out,
                                                                 minutes:dribbles, fixture, opponent_team))
    if (nrow(data.2) == 0) {
      data.2 <- filter(data$history, round == (gameweek-1)) %>%
        select (c(id, kickoff_time, team_h_score:value, selected:transfers_out, minutes:dribbles, fixture, opponent_team))
    }
        
    data.2$element <- i
    points[[i]] <- data.2
    
  }
  
  points <- do.call(rbind, points)
  
  # replace player/team codes with matching values
  points <- merge(points, teams, by.x = "opponent_team", by.y = "id")
  points <- merge(points, teams, by.x = "code", by.y = "code")
  
  # convert player values to fpl-familiar style
  points$price <- points$value/10
  
  # convert var types
  points$influence <- as.numeric(points$influence)
  points$creativity <- as.numeric(points$creativity)
  points$threat <- as.numeric(points$threat)
  points$ict_index <- as.numeric(points$ict_index)
  
  # drop cols
  points <- subset(points, select = -c(opponent_team, value, id.y, code))
  
  # merge player picks with metadata
  picks <- merge(picks, players, by.x = "element", by.y = "id")
  picks <- merge(picks, points, by = "element")
  
  # rename cols
  picks <- dplyr::rename(picks, opponent_team = name.x, player_name = web_name, opponent_team_strength = strength.x,
                         player_team_strength = strength.y, player_team = name.y)
  
  # reorder cols
  picks <- picks[c("round", "fixture", "element", "position", "player_name", "player_team", "player_team_strength",
                   "opponent_team", "opponent_team_strength", "kickoff_time", "team_h_score", "team_a_score", "was_home",
                   "is_captain", "is_vice_captain", "multiplier", "total_points", "price", "selected", "transfers_in", "transfers_out",
                   "minutes", "goals_scored", "assists", "clean_sheets", "goals_conceded", "own_goals", "penalties_saved",
                   "penalties_missed", "yellow_cards", "red_cards", "saves", "bonus", "bps", "influence", "creativity", "threat",
                   "ict_index", "ea_index", "open_play_crosses", "big_chances_created", "clearances_blocks_interceptions",
                   "recoveries", "key_passes", "tackles", "winning_goals", "attempted_passes", "completed_passes", "penalties_conceded",
                   "big_chances_missed", "errors_leading_to_goal", "errors_leading_to_goal_attempt", "tackled", "offside",
                   "target_missed", "fouls", "dribbles")]
  
  return(picks[order(picks$position), ])
  
}