列表列表（来自API调用）到R中的数据框

【问题标题】：List of list of lists (from API call) into data frame in R列表列表（来自API调用）到R中的数据框
【发布时间】：2018-07-18 11:42:30
【问题描述】：

我知道以前有人问过这种性质的问题（将列表转换为数据框），但是我遇到了一个特殊的问题，即我想将其转换为数据框的嵌套列表列表。我得到的数据来自 R 中的 API 调用，因此我要处理这个列表结构的嵌套列表。这是我正在使用的 API 返回对象的一个​​小示例（5 场运动数据）：

dput(soccer_data)
    list(structure(list(id = 1603158L, league_id = 779L, season_id = 914L, 
    stage_id = 1810L, round_id = 29156L, group_id = NULL, aggregate_id = NULL, 
    venue_id = 139L, referee_id = 656L, localteam_id = 607L, 
    visitorteam_id = 3639L, weather_report = NULL, commentaries = TRUE, 
    attendance = NULL, pitch = NULL, winning_odds_calculated = FALSE, 
    formations = structure(list(localteam_formation = "4-2-3-1", 
        visitorteam_formation = "4-1-4-1"), .Names = c("localteam_formation", 
    "visitorteam_formation")), scores = structure(list(localteam_score = 5L, 
        visitorteam_score = 1L, localteam_pen_score = 0L, visitorteam_pen_score = 0L, 
        ht_score = "1-0", ft_score = "5-1", et_score = NULL), .Names = c("localteam_score", 
    "visitorteam_score", "localteam_pen_score", "visitorteam_pen_score", 
    "ht_score", "ft_score", "et_score")), time = structure(list(
        status = "FT", starting_at = structure(list(date_time = "2017-03-04 05:30:00", 
            date = "2017-03-04", time = "05:30:00", timestamp = 1488605400L, 
            timezone = "UTC"), .Names = c("date_time", "date", 
        "time", "timestamp", "timezone")), minute = 90L, extra_minute = NULL, 
        injury_time = NULL), .Names = c("status", "starting_at", 
    "minute", "extra_minute", "injury_time")), coaches = structure(list(
        localteam_coach_id = 429924L, visitorteam_coach_id = 429940L), .Names = c("localteam_coach_id", 
    "visitorteam_coach_id")), standings = structure(list(localteam_position = NULL, 
        visitorteam_position = NULL), .Names = c("localteam_position", 
    "visitorteam_position")), deleted = FALSE), .Names = c("id", 
"league_id", "season_id", "stage_id", "round_id", "group_id", 
"aggregate_id", "venue_id", "referee_id", "localteam_id", "visitorteam_id", 
"weather_report", "commentaries", "attendance", "pitch", "winning_odds_calculated", 
"formations", "scores", "time", "coaches", "standings", "deleted"
)), structure(list(id = 1603159L, league_id = 779L, season_id = 914L, 
    stage_id = 1810L, round_id = 29156L, group_id = NULL, aggregate_id = NULL, 
    venue_id = 113L, referee_id = 3614L, localteam_id = 577L, 
    visitorteam_id = 75L, weather_report = NULL, commentaries = FALSE, 
    attendance = NULL, pitch = NULL, winning_odds_calculated = FALSE, 
    formations = structure(list(localteam_formation = "4-2-3-1", 
        visitorteam_formation = "4-2-3-1"), .Names = c("localteam_formation", 
    "visitorteam_formation")), scores = structure(list(localteam_score = 1L, 
        visitorteam_score = 1L, localteam_pen_score = 0L, visitorteam_pen_score = 0L, 
        ht_score = "1-0", ft_score = "1-1", et_score = NULL), .Names = c("localteam_score", 
    "visitorteam_score", "localteam_pen_score", "visitorteam_pen_score", 
    "ht_score", "ft_score", "et_score")), time = structure(list(
        status = "FT", starting_at = structure(list(date_time = "2017-03-04 22:00:00", 
            date = "2017-03-04", time = "22:00:00", timestamp = 1488664800L, 
            timezone = "UTC"), .Names = c("date_time", "date", 
        "time", "timestamp", "timezone")), minute = 90L, extra_minute = NULL, 
        injury_time = NULL), .Names = c("status", "starting_at", 
    "minute", "extra_minute", "injury_time")), coaches = structure(list(
        localteam_coach_id = 455860L, visitorteam_coach_id = 176760L), .Names = c("localteam_coach_id", 
    "visitorteam_coach_id")), standings = structure(list(localteam_position = NULL, 
        visitorteam_position = NULL), .Names = c("localteam_position", 
    "visitorteam_position")), deleted = FALSE), .Names = c("id", 
"league_id", "season_id", "stage_id", "round_id", "group_id", 
"aggregate_id", "venue_id", "referee_id", "localteam_id", "visitorteam_id", 
"weather_report", "commentaries", "attendance", "pitch", "winning_odds_calculated", 
"formations", "scores", "time", "coaches", "standings", "deleted"
)), structure(list(id = 1603160L, league_id = 779L, season_id = 914L, 
    stage_id = 1810L, round_id = 29156L, group_id = NULL, aggregate_id = NULL, 
    venue_id = 28L, referee_id = 555L, localteam_id = 413L, visitorteam_id = 583L, 
    weather_report = NULL, commentaries = FALSE, attendance = 23554L, 
    pitch = NULL, winning_odds_calculated = FALSE, formations = structure(list(
        localteam_formation = "4-4-1-1", visitorteam_formation = "4-4-2"), .Names = c("localteam_formation", 
    "visitorteam_formation")), scores = structure(list(localteam_score = 1L, 
        visitorteam_score = 2L, localteam_pen_score = 0L, visitorteam_pen_score = 0L, 
        ht_score = "0-0", ft_score = "1-2", et_score = NULL), .Names = c("localteam_score", 
    "visitorteam_score", "localteam_pen_score", "visitorteam_pen_score", 
    "ht_score", "ft_score", "et_score")), time = structure(list(
        status = "FT", starting_at = structure(list(date_time = "2017-03-05 00:00:00", 
            date = "2017-03-05", time = "00:00:00", timestamp = 1488672000L, 
            timezone = "UTC"), .Names = c("date_time", "date", 
        "time", "timestamp", "timezone")), minute = 90L, extra_minute = NULL, 
        injury_time = NULL), .Names = c("status", "starting_at", 
    "minute", "extra_minute", "injury_time")), coaches = structure(list(
        localteam_coach_id = 429914L, visitorteam_coach_id = 429917L), .Names = c("localteam_coach_id", 
    "visitorteam_coach_id")), standings = structure(list(localteam_position = NULL, 
        visitorteam_position = NULL), .Names = c("localteam_position", 
    "visitorteam_position")), deleted = FALSE), .Names = c("id", 
"league_id", "season_id", "stage_id", "round_id", "group_id", 
"aggregate_id", "venue_id", "referee_id", "localteam_id", "visitorteam_id", 
"weather_report", "commentaries", "attendance", "pitch", "winning_odds_calculated", 
"formations", "scores", "time", "coaches", "standings", "deleted"
)), structure(list(id = 1603161L, league_id = 779L, season_id = 914L, 
    stage_id = 1810L, round_id = 29156L, group_id = NULL, aggregate_id = NULL, 
    venue_id = 411L, referee_id = 274L, localteam_id = 1062L, 
    visitorteam_id = 111L, weather_report = NULL, commentaries = FALSE, 
    attendance = NULL, pitch = NULL, winning_odds_calculated = FALSE, 
    formations = structure(list(localteam_formation = "4-2-3-1", 
        visitorteam_formation = "3-5-2"), .Names = c("localteam_formation", 
    "visitorteam_formation")), scores = structure(list(localteam_score = 0L, 
        visitorteam_score = 0L, localteam_pen_score = 0L, visitorteam_pen_score = 0L, 
        ht_score = "0-0", ft_score = "0-0", et_score = NULL), .Names = c("localteam_score", 
    "visitorteam_score", "localteam_pen_score", "visitorteam_pen_score", 
    "ht_score", "ft_score", "et_score")), time = structure(list(
        status = "FT", starting_at = structure(list(date_time = "2017-03-05 00:30:00", 
            date = "2017-03-05", time = "00:30:00", timestamp = 1488673800L, 
            timezone = "UTC"), .Names = c("date_time", "date", 
        "time", "timestamp", "timezone")), minute = 90L, extra_minute = NULL, 
        injury_time = NULL), .Names = c("status", "starting_at", 
    "minute", "extra_minute", "injury_time")), coaches = structure(list(
        localteam_coach_id = 456638L, visitorteam_coach_id = 516577L), .Names = c("localteam_coach_id", 
    "visitorteam_coach_id")), standings = structure(list(localteam_position = NULL, 
        visitorteam_position = NULL), .Names = c("localteam_position", 
    "visitorteam_position")), deleted = FALSE), .Names = c("id", 
"league_id", "season_id", "stage_id", "round_id", "group_id", 
"aggregate_id", "venue_id", "referee_id", "localteam_id", "visitorteam_id", 
"weather_report", "commentaries", "attendance", "pitch", "winning_odds_calculated", 
"formations", "scores", "time", "coaches", "standings", "deleted"
)), structure(list(id = 1603162L, league_id = 779L, season_id = 914L, 
    stage_id = 1810L, round_id = 29157L, group_id = NULL, aggregate_id = NULL, 
    venue_id = 11573L, referee_id = 370L, localteam_id = 179L, 
    visitorteam_id = 641L, weather_report = NULL, commentaries = FALSE, 
    attendance = NULL, pitch = NULL, winning_odds_calculated = FALSE, 
    formations = structure(list(localteam_formation = "4-2-3-1", 
        visitorteam_formation = "4-3-1-2"), .Names = c("localteam_formation", 
    "visitorteam_formation")), scores = structure(list(localteam_score = 1L, 
        visitorteam_score = 0L, localteam_pen_score = 0L, visitorteam_pen_score = 0L, 
        ht_score = "0-0", ft_score = "1-0", et_score = NULL), .Names = c("localteam_score", 
    "visitorteam_score", "localteam_pen_score", "visitorteam_pen_score", 
    "ht_score", "ft_score", "et_score")), time = structure(list(
        status = "FT", starting_at = structure(list(date_time = "2017-03-05 02:00:00", 
            date = "2017-03-05", time = "02:00:00", timestamp = 1488679200L, 
            timezone = "UTC"), .Names = c("date_time", "date", 
        "time", "timestamp", "timezone")), minute = 90L, extra_minute = NULL, 
        injury_time = NULL), .Names = c("status", "starting_at", 
    "minute", "extra_minute", "injury_time")), coaches = structure(list(
        localteam_coach_id = 524071L, visitorteam_coach_id = 261458L), .Names = c("localteam_coach_id", 
    "visitorteam_coach_id")), standings = structure(list(localteam_position = NULL, 
        visitorteam_position = NULL), .Names = c("localteam_position", 
    "visitorteam_position")), deleted = FALSE), .Names = c("id", 
"league_id", "season_id", "stage_id", "round_id", "group_id", 
"aggregate_id", "venue_id", "referee_id", "localteam_id", "visitorteam_id", 
"weather_report", "commentaries", "attendance", "pitch", "winning_odds_calculated", 
"formations", "scores", "time", "coaches", "standings", "deleted"
)))

soccer_data 有 5 场 MLS 足球数据，这是我目前正在做的将其转换为数据框的方法：

# grab the "scores" info from the nested list $scores (from each game)
season_scores <- data.frame()
for(i in 1:length(soccer_data)) {  
  game_scores <- as.data.frame(t(unlist(soccer_data[[i]]$scores)), stringsAsFactors = FALSE)
  game_scores$date <- as.Date(soccer_data[[i]]$time$starting_at$date)
  season_scores <- rbind.fill(season_scores, game_scores)
}
season_scores <- season_scores %>% readr::type_convert()

# create df of the game scores, add the season scores, and drop the bad cols
season_boxscores <- as.data.frame(do.call(rbind, soccer_data), stringsAsFactors = FALSE) %>%
  dplyr::select(-one_of(c('scores', 'group_id', 'aggregate_id', 'time', 'standings'))) %>%
  cbind(season_scores) %>%
  readr::type_convert()

不幸的是，这种方法的问题是最后一个 type_convert() 函数调用没有做我想要的，并且生成的 season_boxscores 数据帧的列的类主要是 class== 列表。

# check yourself
sapply(season_boxscores, class)

那么我的问题是：

  1. 我怎样才能做到这一点，以使 season_boxscores 中所有列的类不都是 class== 列表？还有，
  2. 我是否使用 do.call、rbind 和 as.data.frame 以最佳方式执行此操作（从列表列表转换）？

提前致谢！

编辑：如果所有嵌套列表（在这种情况下，soccer_data 有一些：编队、分数、时间、教练、排名）本身都没有嵌套，就像我取消嵌套它们一样，那就太好了分数的 for 循环。

编辑 2：很抱歉只为 5 场比赛分享这么大的列表对象。在列表列表或像这样的大嵌套对象中，我实际上不知道如何从每个嵌套列表中删除相同的项目，我会为这篇文章做的。 （即从每个soccer_data[[i]] 中删除league_id、round_id 等）。如果有人知道该怎么做，那就太好了！

编辑 3：因为 football_data 不仅仅是一个列表列表，而是一个列表列表（每个列表列表中都有其他非列表对象），这里没有任何解决方案 - Force list of lists into dataframe - 继续足球数据。

【问题讨论】：

    标签： r data-manipulation


    【解决方案1】：

    我还在努力学习这些东西。我测试了一百万件事，这是我能想到的最简单的：

    library(tidyverse)
soccer_data %>% 
  map(unlist) %>% 
  map(t) %>% 
  map(as_tibble) %>% 
  bind_rows()

    想法：获取您的列表soccer_data，将unlist 映射到每个元素（因此它不会在第二级列出，这意味着它将所有游戏保留在最顶层列表的单独元素中）。然后使用 map transpose t 将列表转换为看起来像一行的东西，然后将其转换为tibble，然后将它们全部转换为bind_rows。

    结果：

    # A tibble: 5 x 30
  id      league_id season_id stage_id round_id venue_id referee_id localteam_id
  <chr>   <chr>     <chr>     <chr>    <chr>    <chr>    <chr>      <chr>       
1 1603158 779       914       1810     29156    139      656        607         
2 1603159 779       914       1810     29156    113      3614       577         
3 1603160 779       914       1810     29156    28       555        413         
4 1603161 779       914       1810     29156    411      274        1062        
5 1603162 779       914       1810     29157    11573    370        179         
# ... with 22 more variables: visitorteam_id <chr>, commentaries <chr>,
#   winning_odds_calculated <chr>, formations.localteam_formation <chr>,
#   formations.visitorteam_formation <chr>, scores.localteam_score <chr>,
#   scores.visitorteam_score <chr>, scores.localteam_pen_score <chr>,
#   scores.visitorteam_pen_score <chr>, scores.ht_score <chr>, scores.ft_score <chr>,
#   time.status <chr>, time.starting_at.date_time <chr>, time.starting_at.date <chr>,
#   time.starting_at.time <chr>, time.starting_at.timestamp <chr>,
#   time.starting_at.timezone <chr>, time.minute <chr>,
#   coaches.localteam_coach_id <chr>, coaches.visitorteam_coach_id <chr>,
#   deleted <chr>, attendance <chr>

    看起来对吗？祝你好运！

    【讨论】：

    • 现在试一试，谢谢！ - map 来自哪个 tidyverse 库？编辑：它来自 purrr
    • 当我使用不同的 API 时，我会像每月一次一样重用此代码。再次感谢
    【解决方案2】：

    下面的基本 R 方法怎么样（使用 unlist）：

    1. 将列表列表折叠到list 的char 向量：

      # Collapse list of list of list to list of character vectors
lst <- lapply(soccer_data, unlist);

    2. 确保所有列表条目具有相同的键。例如，只有样本数据的 list 条目 3 具有键 attendance。 

      # Make sure that all list entries have values for the same keys
keys <- unique(unlist(lapply(lst, names)));

    3. 用NA填充缺少的键条目

      # Fill missing entries with NULL
lst <- lapply(lst, function(x) x[match(keys, names(x))]);

    4. rbind 转为data.frame：

      # Combind in dataframe
df <- do.call(rbind.data.frame, lst);
colnames(df) <- keys;


df;
#id league_id season_id stage_id round_id venue_id referee_id
#1 1603158       779       914     1810    29156      139        656
#2 1603159       779       914     1810    29156      113       3614
#3 1603160       779       914     1810    29156       28        555
#4 1603161       779       914     1810    29156      411        274
#5 1603162       779       914     1810    29157    11573        370
#localteam_id visitorteam_id commentaries winning_odds_calculated
#1          607           3639         TRUE                   FALSE
#2          577             75        FALSE                   FALSE
#3          413            583        FALSE                   FALSE
#4         1062            111        FALSE                   FALSE
#5          179            641        FALSE                   FALSE
#formations.localteam_formation formations.visitorteam_formation
#1                        4-2-3-1                          4-1-4-1
#2                        4-2-3-1                          4-2-3-1
#3                        4-4-1-1                            4-4-2
#4                        4-2-3-1                            3-5-2
#5                        4-2-3-1                          4-3-1-2
#scores.localteam_score scores.visitorteam_score scores.localteam_pen_score
#1                      5                        1                          0
#2                      1                        1                          0
#3                      1                        2                          0
#4                      0                        0                          0
#5                      1                        0                          0
#scores.visitorteam_pen_score scores.ht_score scores.ft_score time.status
#1                            0             1-0             5-1          FT
#2                            0             1-0             1-1          FT
#3                            0             0-0             1-2          FT
#4                            0             0-0             0-0          FT
#5                            0             0-0             1-0          FT
#time.starting_at.date_time time.starting_at.date time.starting_at.time
#1        2017-03-04 05:30:00            2017-03-04              05:30:00
#2        2017-03-04 22:00:00            2017-03-04              22:00:00
#3        2017-03-05 00:00:00            2017-03-05              00:00:00
#4        2017-03-05 00:30:00            2017-03-05              00:30:00
#5        2017-03-05 02:00:00            2017-03-05              02:00:00
#time.starting_at.timestamp time.starting_at.timezone time.minute
#1                 1488605400                       UTC          90
#2                 1488664800                       UTC          90
#3                 1488672000                       UTC          90
#4                 1488673800                       UTC          90
#5                 1488679200                       UTC          90
#coaches.localteam_coach_id coaches.visitorteam_coach_id deleted attendance
#1                     429924                       429940   FALSE       <NA>
#2                     455860                       176760   FALSE       <NA>
#3                     429914                       429917   FALSE      23554
#4                     456638                       516577   FALSE       <NA>
#5                     524071                       261458   FALSE       <NA>

    如果你删除所有多余的文字/解释，这很短。

    更新

    很遗憾，unlist 导致列类型丢失。您可以通过以下方式将factors 转换回numeric：

    # Smart-convert to numeric
is.num <- apply(df, 2, function(x) {
    x <- x[!is.na(x)];
    all(suppressWarnings(!is.na(as.numeric(as.character(x)))));
})
df[, is.num] <- apply(df[, is.num], 2, function(x) as.numeric(as.character(x)));

    虽然有点乱，但是很管用。

    【讨论】：

    • 我不喜欢列都是 class== "factor"，因为 readr::type_convert() 无法转换列类型
    • @Canovic 是的，你是对的；我没有意识到所有的列都变成了factors。我已包含将“数字”factors 转换回 numeric 的更新。
    猜你喜欢
    • 2013-02-11
    • 2012-01-18
    • 1970-01-01
    • 2012-09-12
    • 1970-01-01
    • 2021-11-15
    • 1970-01-01
    • 1970-01-01
    • 1970-01-01
    相关资源
    最近更新 更多
    热门标签
    Java Python linux javascript Mysql C# Docker 算法 前端 SpringBoot Redis Vue spring 设计模式 .net core .net kubernetes c++ 数据库 数据结构 大数据 js 机器学习 微服务 Android Go 程序员 面试 JVM ASP.net core 云原生 人工智能 后端 PHP git CSS golang k8s Nginx Django mybatis 深度学习 多线程 React 架构 devops 爬虫 云计算 Spring Boot LeetCode