python

Question

我收到了这个数据集，其中包含 .csv 格式的 key-value 对房地产数据。

如果我删除第一行，我可以用 Pandas 加载它并得到一个像这样的数据框：


id 1	[{'key'": '"floor'"	'"value'": '"2. Floor'"}	{'"key'": '"available_date'"	"value'": '"nach Vereinbarung'"}
id 2	[{'key'": '"floor'"	'"value'": '"1. Floor'"}	{'"key'": '"living_space'"	"value'": 81.0}
id 3	[{'key'": '"living_space'"	'"value'": 240.0}	{'"key'": '"construction_year'"	'"value'": 2012}
id 4	[{'key'": '"living_space'"	'"value'": 280.0}	{'"key'": '"construction_year'"	'"value'": 1851}

但是，我不知道如何使用 Python 中的 key-value 对，所以我想将此数据转换为 Pandas 数据框，其中包含“键”作为每行中的 headers 及其各自的值，如下所示：

id	floor	available_date	living_space	construction_year
id 1	2. Floor		nach Vereinbarung
id 2	1. Floor	81
id 3		240.0		2012
id 4		280.0		1851

我找到了很多关于如何将 Pandas 数据帧转换为 key-value 对的说明，但不是相反...

提前谢谢你。

更新

我的数据内容是这样的：

print(df.head(10))
           [{'key'": '"floor'"   '"value'": '"3. Stock'"}        {'"key'": '"living_space'"    '"value'": 50.0}      {'"key'": '"available_date'"  ... Unnamed: 49 Unnamed: 50 Unnamed: 51 Unnamed: 52 Unnamed: 53
0          [{'key'": '"floor'"   '"value'": '"2. Stock'"}        {'"key'": '"living_space'"   '"value'": 113.0}   {'"key'": '"construction_year'"  ...         NaN         NaN         NaN         NaN         NaN
1          [{'key'": '"floor'"   '"value'": '"1. Stock'"}        {'"key'": '"living_space'"    '"value'": 52.0}   {'"key'": '"construction_year'"  ...         NaN         NaN         NaN         NaN         NaN
..                         ...                        ...                               ...                 ...                               ...  ...         ...         ...         ...         ...         ...
8   [{'key'": '"living_space'"          '"value'": 240.0}   {'"key'": '"construction_year'"    '"value'": 2012}      {'"key'": '"available_date'"  ...         NaN         NaN         NaN         NaN         NaN
9   [{'key'": '"living_space'"          '"value'": 280.0}   {'"key'": '"construction_year'"    '"value'": 1851}      {'"key'": '"available_date'"  ...         NaN         NaN         NaN         NaN         NaN

[10 rows x 54 columns]

更新

.csv 的内容如下所示（第 2 次观察）：

1,"[{'key'"": '""floor'"""," '""value'"": '""3. Stock'""}"," {'""key'"": '""living_space'"""," '""value'"": 50.0}"," {'""key'"": '""available_date'"""," '""value'"": '""01.04.2022'""}"," {'""key'"": '""useful_area'"""," '""value'"": 60.0}"," {'""key'"": '""pets_allowed'"""," '""value'"": true}"," {'""key'"": '""child_friendly'"""," '""value'"": true}"," {'""key'"": '""balcony'"""," '""value'"": true}"," {'""key'"": '""parking_outdoor'"""," '""value'"": true}"," {'""key'"": '""lift'"""," '""value'"": true}"," {'""key'"": '""cable_tv'"""," '""value'"": true}]""","[{'date'"": '""2022-02-25'"""," '""price_amount'"": 1550}]"""

2,"[{'key'"": '""floor'"""," '""value'"": '""2. Stock'""}"," {'""key'"": '""living_space'"""," '""value'"": 113.0}"," {'""key'"": '""construction_year'"""," '""value'"": 2022}"," {'""key'"": '""available_date'"""," '""value'"": '""01.04.2022'""}"," {'""key'"": '""wheelchair_accessible'"""," '""value'"": true}"," {'""key'"": '""child_friendly'"""," '""value'"": true}"," {'""key'"": '""balcony'"""," '""value'"": true}"," {'""key'"": '""parking_indoor'"""," '""value'"": true}"," {'""key'"": '""lift'"""," '""value'"": true}]""","[{'date'"": '""2022-02-27'"""," '""price_amount'"": 2990}]"""

数据似乎是从房地产在线市场中删除的。我认为也与声明每个观察具有不同数量的特征有关。

Answer 1

可能的解决方案如下：

文件'data.csv'内容

1,"[{'key'"": '""floor'"""," '""value'"": '""3. Stock'""}"," {'""key'"": '""living_space'"""," '""value'"": 50.0}"," {'""key'"": '""available_date'"""," '""value'"": '""01.04.2022'""}"," {'""key'"": '""useful_area'"""," '""value'"": 60.0}"," {'""key'"": '""pets_allowed'"""," '""value'"": true}"," {'""key'"": '""child_friendly'"""," '""value'"": true}"," {'""key'"": '""balcony'"""," '""value'"": true}"," {'""key'"": '""parking_outdoor'"""," '""value'"": true}"," {'""key'"": '""lift'"""," '""value'"": true}"," {'""key'"": '""cable_tv'"""," '""value'"": true}]""","[{'date'"": '""2022-02-25'"""," '""price_amount'"": 1550}]"""
2,"[{'key'"": '""floor'"""," '""value'"": '""2. Stock'""}"," {'""key'"": '""living_space'"""," '""value'"": 113.0}"," {'""key'"": '""construction_year'"""," '""value'"": 2022}"," {'""key'"": '""available_date'"""," '""value'"": '""01.04.2022'""}"," {'""key'"": '""wheelchair_accessible'"""," '""value'"": true}"," {'""key'"": '""child_friendly'"""," '""value'"": true}"," {'""key'"": '""balcony'"""," '""value'"": true}"," {'""key'"": '""parking_indoor'"""," '""value'"": true}"," {'""key'"": '""lift'"""," '""value'"": true}]""","[{'date'"": '""2022-02-27'"""," '""price_amount'"": 2990}]"""

import pandas as pd
import json

# read data from csv file
with open("data.csv", "r", encoding="utf-8") as file:
    data = file.read().replace('"', '').replace("'", '"').replace("[", '').replace("]", '').splitlines()

# convert string to list
data_dict = [json.loads("[" + d + "]") for d in data]

data_all = []

for list_item in data_dict:
    data_prepared = {}
    for idx, item in enumerate(list_item):
        if idx == 0:
            data_prepared["id"] = item
        else:
            if 'key' in item:
                data_prepared[item['key']] = item['value']
            else:
                data_prepared.update(item)
    data_all.append(data_prepared)

# create dataframe
df = pd.DataFrame(data_all)
df = df.fillna("-")
df = df.replace(True, 'Yes')
df = df.replace(False, 'No')
df

Returns

python - 如何在 Pandas 数据帧中转换 key-value 对

python - How to transform key-value pairs in Pandas dataframe

key-value

dataframe

pandas