CSV文件的格式必须如何才能使用pandas Dataframe[duplicate]中的列表数据

Question

我的KBs.csv文件：

OS Version,Patch
Server 2019,"['KB4523205', 'KB4519569', 'KB4503308']"

然后我读到：

import pandas as pd
df = pd.read_csv('KBs.csv')
df
OS Version  Patch
0   Server 2019 ['KB4523205', 'KB4519569', 'KB4503308']

令人惊讶的是，当我试图从列表中的任何位置获取数据时，我得到了以下结果：

df['Patch'][0][0]
OUT : '['

如果我尝试构建日期框手册：

dfeo = pd.DataFrame({'OS Version': ['Server 2019'],'Patch': [['KB4523205','KB4519569','KB4503308']]})
dfeo
OS Version  Patch
0   Server 2019 [KB4523205, KB4519569, KB4503308]

dfeo['Patch'][0][0]

OUT : 'KB4523205'

这是我正在寻找的行为，csv文件必须如何格式化才能获得它？

Answer 1

非常感谢你@Chris，是的，你是对的@Trenton这是我要找的“来自ast import literal_eval”

这是完整的代码，它检查了windows sysop上线所需的最小补丁

import pandas as pd
from ast import literal_eval
df = pd.read_csv('KBs.csv', converters={'Patch': literal_eval})

#最小安全补丁2020

w2012 = ['KB4534288', 'KB4537794', 'KB4540694', 'KB4550971', 'KB4556852', 'KB4561674', 'KB4565535']
w2012R2 = ['KB4534309', 'KB4537803', 'KB4541505', 'KB4550970', 'KB4556853', 'KB4561673', 'KB4565540']
w2016 = ['KB4534271', 'KB4537764', 'KB4540670', 'KB4550929', 'KB4556813', 'KB4561616', 'KB4565511']
w2019 = ['KB4561608', 'KB4558998' ]

#“操作系统版本”列的链接

 d = {'Server 2019': w2019, 'Server 2016': w2016, 'Server 2012 R2': w2012R2, 'Server 2012': w2012}

#比较

for k, v in d.items():      
m = df['OS Version'].eq(k)
df.loc[m, 'NoPatch'] =  df.loc[m, 'Patch'].apply(lambda x: list(set(v) - set(x)))
df


Enterprise  Service     Host        OS Version  Patch                    NoPatch
0   Toto    Marketing   consolamig  Server 2019 [KB4523205, KB4519569, KB4503308, KB4499728, K...   [KB4558998, KB4561608]
1   Toto    Marketing   dc01       Server 2016  [KB4525236, KB4485447, KB4520724, KB3192137, K...   [KB4540670, KB4561616, KB4537764, KB4534271, K...
2   Toto    Marketing   duedil01    Server 2012 R2  [KB2920189, KB3103616, KB3045755, KB3161949, K...   [KB4565540, KB4556853, KB4541505, KB4550970, K...
3   Toto    Marketing   win-4ea3    Server 2012 R2  [KB2939471, KB2919355, KB2938772, KB2999226, K...   [KB4565540, KB4556853, KB4541505, KB4550970, K...

再次感谢