R/players.R

In abresler/nbastatR: R's interface to NBA data

.parse_player_json <- function(json, player = player, season = season, mode, measure, season_type) {
  table_length <-
    json$resultSets$rowSet %>% length()
  table_slug <- json$resource

  season_slug <-
    generate_season_slug(season = season)

  all_data <-
    1:table_length %>%
    future_map_dfr(function(x) {
      table_name <-
        json$resultSets$name[x]

      df_parameters <- json$parameters %>% flatten_df()

      df_parameters <-
        df_parameters %>%
        set_names(names(df_parameters) %>% resolve_nba_names()) %>%
        munge_nba_data() %>%
        select(-matches("remove"))
      df_parameters <-
        df_parameters %>%
        mutate_at(df_parameters %>% dplyr::select(dplyr::matches("is[A-Z]")) %>% names(),
                  funs(ifelse(. == "Y", 1, 0) %>% as.logical())) %>%
        mutate(numberTable = x,) %>%
        select(one_of(c("numberTable", "typeMeasure", "modeSearch")), everything()) %>%
        suppressWarnings()

      json_names <-
        json$resultSets$headers[[x]]

      actual_names <-
        json_names %>%
        resolve_nba_names()

      data <-
        json$resultSets$rowSet[[x]] %>%
        data.frame(stringsAsFactors = F) %>%
        dplyr::as_tibble() %>%
        select(-matches("remove"))

      if (data %>% nrow() == 0) {
        return(invisible())
      }

      data <-
        data  %>%
        set_names(actual_names) %>%
        munge_nba_data() %>%
        mutate(numberTable = x,
               namePlayer = player,
               slugSeason = season_slug,
               typeMeasure = measure,
               modeSearch = mode,
               typeSeason = season_type) %>%
        dplyr::select(one_of(
          c(
            "typeSeason",
            "modeSearch",
            "typeMeasure",
            "numberTable",
            "slugSeason",
            "namePlayer"
          )
        ), everything()) %>%
        suppressWarnings()

      if (table_name == "NextNGames") {
        data <-
          data %>%
          unite(nameTeamHome, cityTeamHome, teamNameHome,  sep =  " ") %>%
          unite(nameTeamAway, cityTeamAway, teamNameAway,  sep =  " ") %>%
          mutate(dateGame = dateGame %>% lubridate::mdy())
      }

      if (data %>% has_name("typeShot")){
        data <-
          data %>%
          dplyr::rename(typeFilter = typeShot)
      }

      if (table_name == "PlayersSeasonTotals"){
        if (data %>% has_name("namePlayer")) {
          data <-
            data %>%
            dplyr::rename(typeFilter = namePlayer)
        }

      }

      if (table_name == "TeamOverall"){
      data <-
        data %>%
          mutate(nameGroup = "Players")

      }
      if (table_name == "OverallPlayerDashboard") {
        return(invisible())
      }

      if (table_name %in% c("OverallPlayerDashboard", "ByYearPlayerDashboard")){
        data <-
          data %>%
          dplyr::rename(slugSeasonSearch = slugSeason,
                        slugSeason = typeFilter) %>%
          mutate(modeSearch = mode,
                 typeMeasure = measure,
                 namePlayer = player)

      }

      data <-
        data %>%
        left_join(df_parameters) %>%
        dplyr::select(one_of(names(df_parameters)), everything()) %>%
        suppressMessages() %>%
        select(-numberTable) %>%
        mutate(nameTable = table_name) %>%
        select(nameTable, everything()) %>%
        dplyr::select(-one_of("idLeague")) %>%
        remove_zero_sum_cols() %>%
        mutate(slugTable = table_slug,
               yearSeason = season) %>%
        suppressWarnings() %>%
        suppressMessages()

      data <-
        data %>%
        dplyr::select(nameTable:slugSeason, yearSeason, everything())

      data <-
        data %>%
        dplyr::select(-one_of(c( "resultGame", "locationGame",
                                 "nameSeasonMonth", "segmentSeason", "rangeDaysRest"))) %>%
        suppressWarnings()

      if (table_name == "ByYearTeamDashboard") {
        if (data %>% has_name("slugSeason")) {
        data <-
          data %>%
          dplyr::rename(slugSeasonSearch = slugSeason)
        }

        if (data %>% has_name("groupByYear")) {
          data <-
            data %>%
            dplyr::rename(slugSeason = groupByYear)
        }
      }


      if (table_name %in% c("OverallTeamDashboard", "OverallTeamPlayerOnOffSummary")) {
        return(invisible())
      }

      if (table_name == "AssistedBy") {
        assign_nba_players()
        data <-
          data %>%
          dplyr::rename(idPlayerAssistedBy = idPlayer) %>%
          dplyr::select(-one_of("namePlayer")) %>%
          left_join(
            df_dict_nba_players %>% select(idPlayerAssistedBy = idPlayer,
                                           namePlayerAssistedBy = namePlayer)
          ) %>%
          suppressMessages()

        data <-
          data %>%
          dplyr::select(dplyr::matches("type|mode|^is|^id|^name"),
                        everything())
      }

      key_cols <-
        c("slugTable", "nameTable","typeSeason", "slugSeason", "yearSeason", "slugSeasonSearch", "namePlayer",
          "typeMeasure", "modeSearch") %>% unique()

      nest_cols <-
        names(data)[!names(data) %in% key_cols]

      data <-
        data %>%
        nest(.key = "dataTable",cols = nest_cols) %>%
        rename(dataTable = cols)

      data
    })

  all_data

}

.dictionary_player_tables <-
  memoise(function() {
    tibble(
      nameTable = c(
        "passes",
        "clutch",
        "game splits",
        "general splits",
        "opponent",
        "next n games",
        "player on off details",
        "defense",
        "game logs",
        "rebounding",
        "shot chart detail",
        "shots",
        "year over year",
        "fantasy profile"

      ),
      slugTable = c(
        "playerdashptpass",
        "playerdashboardbyclutch",
        "playerdashboardbygamesplits",
        "playerdashboardbygeneralsplits",
        "playerdashboardbyopponent",
        "playernextngames",
        "leagueplayerondetails",
        "playerdashptshotdefend",
        "playergamelogs",
        "playerdashptreb",
        "shotchartdetail",
        "playerdashptshots",
        "playerdashboardbyyearoveryear",
        "playerfantasyprofile"
      )
    )
  })

.get_player_table_data <-
  function(player_id = 1627747,
           table = "year over year",
           measure = "Base",
           season = 2018,
           mode = "PerGame",
           season_type = "Regular Season",
           game_id = NA,
           n_game = NA,
           context_measure = "FGM",
           playoff_round = NA,
           is_plus_minus = F,
           is_rank = F,
           is_pace_adjusted = F,
           outcome = NA,
           location = NA,
           month = NA,
           season_segment = NA,
           date_from = NA,
           date_to = NA,
           opponent_id = NA,
           vs_conf = NA,
           vs_division = NA,
           game_segment = NA,
           period = NA,
           shot_clock =  NA,
           last_n_games = NA,
           return_message = TRUE) {
    df_player_slug_tables <-
      .dictionary_player_tables()
    assign_nba_players()

    player <-
      df_dict_nba_players %>%
      filter(idPlayer == player_id) %>%
      pull(namePlayer)

    if (return_message) {
      glue("Acquiring {player} {season} {season_type} {measure} {table} {mode} data") %>% cat(fill = T)
    }

    table_slug <-
      df_player_slug_tables %>%
      filter(nameTable  == (str_to_lower(table))) %>%
      pull(slugTable)
    URL <- gen_url(table_slug)
    measure_slug <-
      generate_call_slug(x = str_to_title(measure), default_value = "Base")
    mode_slug <-
      generate_call_slug(x = mode, default_value = "PerGame")
    context_measure_slug = generate_call_slug(x = context_measure, default_value = "")
    season_slug <- generate_season_slug(season = season)
    game_id_slug <-
      generate_call_slug(x = game_id, default_value = 0)
    season_type_slug  = generate_call_slug(x = season_type, default_value = "Regular+Season")
    playoff_round_slug = generate_call_slug(x = playoff_round, default_value = 0)
    plus_minus_slug <-
      generate_call_slug(x = is_plus_minus , default_value = "N")
    rank_slug <-
      generate_call_slug(x = is_rank , default_value = "N")
    pace_slug <-
      generate_call_slug(x = is_pace_adjusted , default_value = "N")
    outcome_slug <-
      generate_call_slug(x = outcome , default_value = "")
    location_slug <-
      generate_call_slug(x = location , default_value = "")
    month_slug <- generate_call_slug(x = month , default_value = 0)
    season_segment_slug <-
      generate_call_slug(x = season_segment , default_value = "")
    date_from_slug <-
      generate_call_slug(x = date_from , default_value = "")
    date_to_slug <-
      generate_call_slug(x = date_to , default_value = "")
    opponent_id_slug <-
      generate_call_slug(x = opponent_id , default_value = 0)
    vs_conf_slug <-
      generate_call_slug(x = season_segment , default_value = "")
    vs_division_slug <-
      generate_call_slug(x = vs_division , default_value = "")
    game_segment_slug  <-
      generate_call_slug(x = game_segment , default_value = "")
    period_slug <-
      generate_call_slug(x = period , default_value = 0)
    shot_clock_slug <-
      generate_call_slug(x = shot_clock , default_value = "")
    last_n_games_slug <-
      generate_call_slug(x = last_n_games , default_value = 0)
    game_number_slug <-
      generate_call_slug(x = n_game , default_value = 0)
    params <-
      list(
        measureType = measure_slug,
        perMode = mode_slug,
        plusMinus = plus_minus_slug,
        contextMeasure = context_measure_slug,
        paceAdjust = pace_slug,
        rank = rank_slug,
        leagueId = "00",
        season = season_slug,
        seasonType = season_type,
        GameID = game_id_slug,
        TeamID = 0,
        GROUP_ID = 0,
        numberOfGames = game_number_slug,
        poRound = playoff_round_slug,
        playerID = player_id,
        outcome = outcome_slug,
        location = location_slug,
        month = month_slug,
        seasonSegment = season_segment_slug,
        dateFrom = date_from_slug,
        dateTo = date_to_slug,
        opponentTeamId = opponent_id_slug,
        vsConference = vs_conf_slug,
        vsDivision = vs_division_slug,
        gameSegment = game_segment_slug,
        period = period_slug,
        shotClockRange = shot_clock_slug,
        lastNGames = last_n_games_slug
      )
    if (table_slug == "teamvsplayer") {
      names(params)[names(params) %>% str_detect("teamId")] <-
        "playerId"
    }
    slug_param <-
      .generate_param_slug(params = params)

    url <-
      glue("{URL}?{slug_param}") %>% as.character()

    json <-
      url %>%
      .curl_chinazi(
      )

    all_data <-
      .parse_player_json(
        json = json,
        player = player,
        season = season,
        mode = mode,
        measure = measure,
        season_type = season_type
      ) %>%
      mutate(idPlayer = player_id,
             typeMeasure = measure,
             modeSearch = mode,
             slugSeason = season_slug,
             yearSeason = season)  %>%
      dplyr::select(one_of(c("nameTable", "typeMeasure", "modeSearch", "slugSeason", "yearSeason",
                             "typeSeason", "slugSeasonSearch",
                             "idPlayer", "namePlayer", "slugTable",
                             "dataTable")
      ), everything()) %>%
      suppressWarnings()
    all_data
  }


#' NBA players table data
#'
#' @param players vector of player names
#' @param player_ids vector of player ids
#' @param tables vector of tables options \itemize{
#' \item passes
#' \item clutch
#' \item game splits
#' \item general splits
#' \item opponent
#' \item next n games
#' \item player on off details
#' \item defense
#' \item game logs
#' \item rebounding
#' \item shot chart detail
#' \item shots
#' \item year over year
#' \item fantasy profile
#' }
#' @param measures vector of measure types options include \itemize{
#' \item Base
#' \item Advanced
#' \item Misc
#' \item Scoring
#' \item Four Factors
#' \item Opponent
#' \item Usage
#' \item Defense
#' }
#' @param seasons vector of seasons
#' @param modes vector of modes options include \itemize{
#' \item PerGame
#' \item Totals
#' \item MinutesPer
#' \item Per48
#' \item Per40
#' \item Per36
#' \item PerMinute
#' \item PerPossession
#' \item PerPlay
#' \item Per100Possessions
#' \item Per100Plays
#' }#'
#' @param playoff_rounds vector of playoff rounds options include code{0:4}
#' @param is_plus_minus \code{TRUE} returns plus minus
#' @param is_rank if \code{TRUE} returns rank
#' @param is_pace_adjusted if \code{TRUE} adjusts for pace
#' @param outcomes vector of outcomes options include \itemize{
#' \item NA
#' \item Wins
#' \item Losses
#' }
#' @param locations vector of locations options include \itemize{
#' \item NA
#' \item Home
#' \item Road
#' }
#' @param months vector of game months options include \code{0:12}
#' @param season_segments vector of season segments, options include \itemize{
#' \item NA
#' \item Post All-Star
#' \item Pre All-Star
#' }
#' @param date_from \code{NA} or date from
#' @param date_to \code{NA} or date to
#' @param opponent_ids vector of opponent ids
#' @param vs_confs vector of conferences against options include  \itemize{
#' \item NA
#' \item East
#' \item West
#' }
#' @param vs_divisions vector of divisions against options include \itemize{
#' \item NA
#' \item Atlantic
#' \item Central
#' \item Northwest
#' \item Pacific
#' \item Southeast
#' \item Southwest
#' }
#' @param game_segments vector of game segments options include \itemize{
#' \item NA
#' \item First Half
#' \item Second Half
#' \item Overtime
#' }
#' @param periods vector of periods \code{0:12}
#' @param last_n_games vector of last_n games \code{0:82}
#' @param season_types vector of season types options include \itemize{
#' \item Regular Season
#' \item Pre Season
#' \item Playoffs
#' \item All Star
#' }
#' @param n_games number n of last games
#' @param shot_clocks  vector of shot clock ranges options include \itemize{
#' \item  NA,
#' \item 24-22
#' \item 22-18 Very Early
#' \item 18-15 Early
#' \item 15-7 Average
#' \item 7-4 Late
#' \item 4-0 Very Late
#' \item ShotClock Off
#' }
#' @param return_message if \code{TRUE} returns a message
#' @param assign_to_environment if \code{TRUE} assigns data to environment
#'
#' @return a \code{tibble}
#'
#' @export
#' @family player
#' @examples
#' players_tables(players = c("Caris LeVert", "Joe Harris"), tables =  c("year over year", "passes", "game splits"),   modes = c("PerGame", "Totals"), measures = c("Base", "Advanced"), assign_to_environment = TRUE)
players_tables <-
  function(players = NULL,
           player_ids = NULL,
           tables = c("year over year", "passes", "game splits"),
           measures = "Base",
           seasons = 2019,
           modes = c("PerGame", "Totals"),
           season_types = "Regular Season",
           playoff_rounds = NA,
           is_plus_minus = F,
           n_games = 20,
           is_rank = F,
           is_pace_adjusted = F,
           outcomes = NA,
           locations = NA,
           months = NA,
           season_segments = NA,
           date_from = NA,
           date_to = NA,
           opponent_ids = NA,
           vs_confs = NA,
           vs_divisions = NA,
           game_segments = NA,
           periods = NA,
           shot_clocks =  NA,
           last_n_games = NA,
           assign_to_environment = TRUE,
           return_message = TRUE) {

    assign_nba_players()

    ids <-
      nba_player_ids(player_ids = player_ids,
                          players = players)

    input_df <-
      expand.grid(
        player_id = ids,
        table = tables,
        measure = measures,
        season = seasons,
        mode = modes,
        n_game = n_games,
        season_type = season_types,
        playoff_round = playoff_rounds,
        is_plus_minus = is_plus_minus,
        is_rank = is_rank,
        is_pace_adjusted = is_pace_adjusted,
        outcome =  outcomes,
        location = locations,
        month  = months,
        season_segment = season_segments,
        date_from = date_from,
        date_to = date_to,
        opponent_id = opponent_ids,
        vs_conf = vs_confs,
        vs_division = vs_divisions,
        game_segment = game_segments,
        period = periods,
        shot_clock = shot_clocks,
        last_n_games = last_n_games,
        stringsAsFactors = F
      ) %>%
      dplyr::as_tibble()
    get_player_table_data_safe <-
      possibly(.get_player_table_data, tibble())

    all_data <-
      1:nrow(input_df) %>%
      future_map_dfr(function(x) {
        df_row <-
          input_df %>% slice(x)
        df_row %$%
          get_player_table_data_safe(
            player_id = player_id,
            table = table,
            measure = measure,
            season = season,
            mode = mode,
            season_type = season_type,
            game_id = NA,
            context_measure = NA,
            playoff_round = playoff_round,
            is_plus_minus = is_plus_minus,
            is_rank = is_rank,
            is_pace_adjusted = is_pace_adjusted,
            outcome = outcome,
            location = location,
            month = month,
            season_segment = season_segment,
            date_from = date_from,
            date_to = date_to,
            opponent_id = opponent_id,
            vs_conf = vs_conf,
            vs_division = vs_division,
            game_segment = game_segment,
            period = period,
            shot_clock = shot_clock,
            last_n_games = last_n_games,
            return_message = return_message
          )
      })
    df_dict_table_names <-
      .dictionary_player_tables()

    table_names <-
      df_dict_table_names$nameTable %>% map_chr(function(x) {
        generate_data_name(x = x, result = "Player")
      })

    df_dict_table_names <-
      df_dict_table_names %>%
      mutate(tableName = table_names) %>%
      select(-nameTable) %>%
      dplyr::rename(tableSlugName = tableName)

    all_data <-
      all_data %>%
      left_join(df_dict_nba_players %>% select(idPlayer, dplyr::matches("url"))) %>%
      suppressWarnings() %>%
      suppressMessages()

    all_data <-
      all_data %>%
      left_join(df_dict_table_names) %>%
      select(tableSlugName, nameTable, everything()) %>%
      suppressMessages() %>%
      unique()

    if (assign_to_environment) {
      all_tables <-
        all_data$tableSlugName %>%
        unique()
      all_tables %>%
        walk(function(table) {
          df_tables <-
            all_data %>%
            filter(tableSlugName == table) %>%
            select(-one_of(c("slugTable", "tableSlugName"))) %>%
            unnest() %>%
            remove_na_columns()


          has_measure <- df_tables %>% has_name("typeMeasure")

          if (has_measure) {
            measures <-
              df_tables$typeMeasure %>% unique()
            measures %>%
              walk(function(measure) {
                table_name <-
                  table %>%
                  str_c(measure)
                df_table <-
                  df_tables %>%
                  filter(typeMeasure == measure) %>%
                  unnest() %>%
                  remove_na_columns() %>%
                  distinct()
                assign(x = table_name,
                       value = df_table,
                       envir = .GlobalEnv)
              })
          } else{
            df_table <-
              df_tables %>%
              unnest() %>%
              remove_na_columns() %>%
              distinct()

            if (df_table %>% has_name("idPlayer1")){
              data <- data %>% select(-idPlayer1)
            }

            assign(x = table,
                   value = df_table,
                   envir = .GlobalEnv)
          }
        })
    }
    all_data %>%
      remove_na_columns()
  }