从python中的单个csv文件创建嵌套词典列表

Question

我有一个具有以下结构的csv文件：

team,tournament,player
Team 1,spring tournament,Rebbecca Cardone
Team 1,spring tournament,Salina Youngblood
Team 1,spring tournament,Catarina Corbell
Team 1,summer tournament,Cara Mejias
Team 1,summer tournament,Catarina Corbell
...
Team 10, spring tournament,Jessi Ravelo

我想创建一个嵌套字典（团队、锦标赛），其中包含玩家字典列表。期望的结果是：

{'data': 
   {Team 1:
       {'spring tournament': 
               {'players': [
                   {name: Rebecca Cardone}, 
                   {name: Salina Youngblood},
                   {name: Catarina Corbell}]
        },
        {'summer tournament': 
               {'players': [
                   {name: Cara Mejias}, 
                   {name: Catarina Corbell}]
               }
        }

   },
    ...
  {Team 10:
       {'spring tournament': 
               {'players': [
                   {name: Jessi Ravelo}]
                }
       }
   }
}

我一直在努力把它格式化成这样。我已经能够成功地嵌套第一级（团队-->；锦标赛），但我无法嵌套第二级。目前，我的代码如下所示：

d = {}
header = True
with open("input.csv") as f:
    for line in f.readlines():
        if header:
            header = False
            continue
        team, tournament, player = line.strip().split(",")
        d_team = d.get(team,{})
        d_tournament = d_team.get(tournament, {})
        d_player = d_tournament.get('player',['name'])
        d_player.append(player)
        d_tournament['player'] = d_tournament
        d_team[tournament] = d_tournament
        d[team] = d_team
print(d)

修复代码以便创建嵌套字典的下一步是什么

Answer 1

您描述的示例字典是不可能的（如果您希望在键“Team 1”下有多个字典，请将它们放在列表中），但此代码段：

if __name__ == '__main__':
    your_dict = {}
    with open("yourfile.csv") as file:
        all_lines = file.readlines()

    data_lines = all_lines[1:]  #  Skipping "team,tournament,player" line

    for line in data_lines:
        line = line.strip()  # Remove \n
        team, tournament_type, player_name = line.split(",")
        team_dict = your_dict.get(team, {})  # e.g. "Team 1"

        tournaments_of_team_dict = team_dict.get(tournament_type, {'players': []})  # e.g. "spring_tournament"

        tournaments_of_team_dict["players"].append({'name': player_name})

        team_dict[tournament_type] = tournaments_of_team_dict
        your_dict[team] = team_dict

    your_dict = {'data': your_dict}

对于本例yourfile.csv：

team,tournament,player
Team 1,spring tournament,Rebbecca Cardone
Team 1,spring tournament,Salina Youngblood
Team 2,spring tournament,Catarina Corbell
Team 1,summer tournament,Cara Mejias
Team 2,summer tournament,Catarina Corbell

提供以下信息：

{
  "data": {
    "Team 1": {
      "spring tournament": {
        "players": [
          {
            "name": "Rebbecca Cardone"
          },
          {
            "name": "Salina Youngblood"
          }
        ]
      },
      "summer tournament": {
        "players": [
          {
            "name": "Cara Mejias"
          }
        ]
      }
    },
    "Team 2": {
      "spring tournament": {
        "players": [
          {
            "name": "Catarina Corbell"
          }
        ]
      },
      "summer tournament": {
        "players": [
          {
            "name": "Catarina Corbell"
          }
        ]
      }
    }
  }
}

Process finished with exit code 0

Answer 2

您的实现中存在一些问题：

• 你可以d_player = d_tournament.get('player',['name'])。但实际上，您希望获得名为players的键，这应该是一个字典列表。这些词典的格式必须为{"name": "Player's Name"}。所以你想要 l_player = d_tournament.get('players',[])（默认为空列表），然后执行l_player.append({"name": player})（我将其重命名为l_player，因为它是一个列表，而不是dict）
• 你可以d_tournament['player'] = d_tournament。我猜你的意思是d_tournament['player'] = d_player
• 去除行中元素的空白。Doteam, tournament, player = (word.strip() for word in line.split(","))

在进行这些更改后，代码可以正常工作

---

我强烈建议您使用csv.reader类来读取CSV文件，而不是手动用逗号分隔行

此外，由于python的容器（列表和字典）包含对其内容的引用，您只需添加容器一次，然后使用mydict["key"] = value或mylist.append()对其进行修改，这些更改也将反映在父容器中。由于这种行为，您不需要像使用d_team[tournament] = d_tournament一样在循环中重复分配这些内容

allteams = dict()
hasHeader = True
with open("input.csv") as f:
    csvreader = csv.reader(f)
    if hasHeader: next(csvreader) # Consume one line if a header exists

    # Iterate over the rows, and unpack each row into three variables
    for team_name, tournament_name, player_name in csvreader:
        # If the team hasn't been processed yet, create a new dict for it
        if team_name not in allteams:
            allteams[team_name] = dict()

        # Get the dict object that holds this team's information
        team = allteams[team_name]

        # If the tournament hasn't been processed already for this team, create a new dict for it in the team's dict
        if tournament_name not in team:
            team[tournament_name] = {"players": []}

        # Get the tournament dict object
        tournament = team[tournament_name]

        # Add this player's information to the tournament dict's "player" list
        tournament["players"].append({"name": player_name})

# Add all teams' data to the "data" key in our result dict
result = {"data": allteams}
print(result)    

这给了我们想要的（美化输出）：

{
    'data': {
        'Team 1': {
            'spring tournament': {
                'players': [
                    { 'name': 'Rebbecca Cardone' },
                    { 'name': 'Salina Youngblood' },
                    { 'name': 'Catarina Corbell' }
                ]
            },
            'summer tournament': {
                'players': [
                    { 'name': 'Cara Mejias' },
                    { 'name': 'Catarina Corbell' }
                ]
            }
        },
        'Team 10': {
            ' spring tournament': {
                'players': [
                    { 'name': 'Jessi Ravelo' }
                ]
            }
        }
    }
}

Answer 3

也许我忽略了什么，但你不能用：

df.groupby(['team','tournament'])['player'].apply(list).reset_index().to_json(orient='records')