使用 python 构建树

Question

我是 python 的新手，并尝试使用 python 构建具有父子关系的树。 我很难做到：

数据框：

data = [['root','Parent1','Children1','Grand Childern 1','Great Grand Childern 1'],['root','Parent1','Children2','Grand Childern 1','Great Grand Childern 1'],['root','Parent1','Children2','Grand Childern 2','Great Grand Childern 1'],['root','Parent2','Children1','Grand Childern 1','Great Grand Childern 1'],['root','Parent2','Children2','Grand Childern 1','Great Grand Childern 1']]

df=pd.DataFrame(data,columns=['LEVEL 1','LEVEL 2','LEVEL 3','LEVEL 4','LEVEL 5'])

我正在尝试转换为 JSON 树格式，如下所示：

{
"name": "root",
"children": [{ 
    "name": "Parent1",
    "children": [{ 
        "name": "Children1" ,
        "children":[{
            "name":"Grand Children1",
            "children":[{
                "name":"Great Grand Children1"
                }]
            }]},
    { 
        "name": "Children2" , 
        "children":[{
          "name":"Grand Children1",
                "children":[{"name":"Great Grand Children1"}],
          "name":"Grand Children2",
                "children":[{"name":"Great Grand Children1"}]}
          ] }
    ]
  },
  { 
    "name": "Parent2",
    "children": [
      { 
          "name": "Children1" ,
          "children":[{"name":"Grand Children1",
                        "children":[{"name":"Great Grand Children1"}]}] },
      { 
          "name": "Children2" , 
          "children":[{
          "name":"Grand Children1","children":[{"name":"Great Grand Children1"}],
          "name":"Grand Children1","children":[{"name":"Great Grand Children2"}]}
          ] }
    ]
  }]
}

DataFrame：

如果有人可以帮助我，我将不胜感激。

Answer 1

作为中间步骤，您可能希望将列表列表（当前不在树结构中，而是指定树的各个垂直分支）转换为字典树，其中每个节点都包含对其所有节点的引用子节点。

使用 dict 从列表列表中构建树可以非常简单地确保任何给定节点的所有子节点都位于正确的位置（即在该节点下组合在一起）：

>>> data = [
    ['root', 'Parent1', 'Children1', 'Grand Childern 1', 'Great Grand Childern 1'],
    ['root', 'Parent1', 'Children2', 'Grand Childern 1', 'Great Grand Childern 1'],
    ['root', 'Parent1', 'Children2', 'Grand Childern 2', 'Great Grand Childern 1'],
    ['root', 'Parent2', 'Children1', 'Grand Childern 1', 'Great Grand Childern 1'],
    ['root', 'Parent2', 'Children2', 'Grand Childern 1', 'Great Grand Childern 1']
]
>>> tree = {}
>>> for row in data:
...     node = tree
...     for cell in row:
...         node = node.setdefault(cell, {})
...
>>> tree
{'root': {
    'Parent1': {
        'Children1': {
            'Grand Childern 1': {
                'Great Grand Childern 1': {}
            }
        }, 
        'Children2': {
            'Grand Childern 1': {
                'Great Grand Childern 1': {}
            }, 
            'Grand Childern 2': {
                'Great Grand Childern 1': {}
            }
        }
    }, 
    'Parent2': {
        'Children1': {
            'Grand Childern 1': {
                'Great Grand Childern 1': {}
            }
        }, 
        'Children2': {
            'Grand Childern 1': {
                'Great Grand Childern 1': {}
            }
        }
    }
}}

现在您已经将所有内容都放在了一个实际的树结构中，将其转换为您需要的任何更具体的格式（例如所需的 JSON）应该很简单。

Answer 2

如果它不必来自“Pandas”库，您可以使用“ anytree ”库创建树，并使用“ JSON Exporter ”将其导出到 JSON。

Answer 3

您可以将递归与collections.defaultdict一起使用：

from collections import defaultdict
def to_tree(d):
   _d = defaultdict(list)
   for a, *b in d:
      _d[a].append(b)
   return [{'name':a, 'children':to_tree(k)} if (k:=list(filter(None, b))) else \
                    {'name':a} for a, b in _d.items()]

data = [['root','Parent1','Children1','Grand Childern 1','Great Grand Childern 1'],['root','Parent1','Children2','Grand Childern 1','Great Grand Childern 1'],['root','Parent1','Children2','Grand Childern 2','Great Grand Childern 1'],['root','Parent2','Children1','Grand Childern 1','Great Grand Childern 1'],['root','Parent2','Children2','Grand Childern 1','Great Grand Childern 1']]

---

import json
print(json.dumps(to_tree(data), indent=4))

Output：

[
  {
    "name": "root",
    "children": [
        {
            "name": "Parent1",
            "children": [
                {
                    "name": "Children1",
                    "children": [
                        {
                            "name": "Grand Childern 1",
                            "children": [
                                {
                                    "name": "Great Grand Childern 1"
                                }
                            ]
                        }
                    ]
                },
                {
                    "name": "Children2",
                    "children": [
                        {
                            "name": "Grand Childern 1",
                            "children": [
                                {
                                    "name": "Great Grand Childern 1"
                                }
                            ]
                        },
                        {
                            "name": "Grand Childern 2",
                            "children": [
                                {
                                    "name": "Great Grand Childern 1"
                                }
                            ]
                        }
                    ]
                }
            ]
        },
        {
            "name": "Parent2",
            "children": [
                {
                    "name": "Children1",
                    "children": [
                        {
                            "name": "Grand Childern 1",
                            "children": [
                                {
                                    "name": "Great Grand Childern 1"
                                }
                            ]
                        }
                    ]
                },
                {
                    "name": "Children2",
                    "children": [
                        {
                            "name": "Grand Childern 1",
                            "children": [
                                {
                                    "name": "Great Grand Childern 1"
                                }
                            ]
                        }
                    ]
                 }
             ]
          }
      ]
   }
]

没有Python3.8赋值表达式的解决方案：

from collections import defaultdict
def to_tree(d):
   _d = defaultdict(list)
   for a, *b in d:
     _d[a].append(b)
   vals = [[a, list(filter(None, b))] for a, b in _d.items()]
   return [{'name':a, 'children':to_tree(b)} if b else {'name':a} for a, b in vals]