从熊猫列中提取特定值并将其存储在新列中

Question

我有一个熊猫栏，它以下列格式以列表的形式存储数据：

text
[['Mark','PERSON'],['Data Scientist','TITLE'], ['Berlin','LOC'], ['Python','SKILLS'], ['Tableau,','SKILLS'], ['SQL','SKILLS'], ['AWS','SKILLS']]
[['John','PERSON'],['Data Engineer','TITLE'], ['London','LOC'], ['Python','SKILLS'], ['DB2,','SKILLS'], ['SQL','SKILLS']
[['Pearson','PERSON'],['Intern','TITLE'], ['Barcelona','LOC'], ['Python','SKILLS'], ['Excel,','SKILLS'], ['SQL','SKILLS']
[['Broody','PERSON'],['Manager','TITLE'], ['Barcelona','LOC'], ['Team Management','SKILLS'], ['Excel,','SKILLS'], ['Good Communications','SKILLS']
[['Rita','PERSON'],['Software Developer','TITLE'], ['London','LOC'], ['Dot Net','SKILLS'], ['SQl Server,','SKILLS'], ['VS Code,'SKILLS']

作为输出，我想看到的是：

PERSON    TITLE                 LOC         SKILLS       
Mark      Data Scientist        Berlin      Python, Tableau, SQL, AWS
John      Data Engineer         London      Python, DB2,SQL

.其他输入行也是如此。

因此，从本质上说，将数据拆分为"，“并将左边部分存储在"，”作为列标题，而将“右边的部分”存储为值。

我怎样才能做到这一点？

Answer 1

如果您有一个像这样的数据框架叫做"df":，那么

   index    text
0    1     [[Mark, PERSON], [Data Scientist, TITLE], [Ber...
1    2     [[John, PERSON], [Data Engineer, TITLE], [Lond...
2    3     [[Pearson, PERSON], [Intern, TITLE], [Barcelon...
3    4     [[Broody, PERSON], [Manager, TITLE], [Barcelon...
4    5     [[Rita, PERSON], [Software Developer, TITLE], ...

• 你可以试试这样的东西:

person=[]
skills=[]
title=[]
loc=[]
temp=[]

for i in range(len(df['text'])):
    
    for j in range(len(df['text'][i])):
        
        if df['text'][i][j][1]=='PERSON':
            
            person.append(df['text'][i][j][0])
        
        elif df['text'][i][j][1]=='TITLE':
            
            title.append(df['text'][i][j][0])
        
        elif df['text'][i][j][1]=='LOC':
            
            loc.append(df['text'][i][j][0])
        
        elif df['text'][i][j][1]=='SKILLS':
            
            temp.append(df['text'][i][j][0].replace(",", ""))
    
    skills.append(",".join(temp))
    temp=[]

• Output

    PERSON  TITLE              LOC           SKILLS
0   Mark    Data Scientist    Berlin    Python,Tableau,SQL,AWS
1   John    Data Engineer     London    Python,DB2,SQL
2   Pearson Intern            Barcelona Python,Excel,SQL
3   Broody  Manager           Barcelona Team Management,Excel,Good Communications
4   Rita    Software Developer London   Dot Net,SQl Server,VS Code