将元素中的列表转换为字典或替换Python中pandas数据框中元素中的特殊字符

Question

I'm working on a dataframe like this in pandas:

item_type  item_price  item_env                                        item_detail
sofa          2880     [natural wood, eco-friendly]                 [2pax,seglora]     
sofa          2400     [solid wood, recycle, reuse]                 [3pax,knisa]
chair         1200      nan                                               [Red]
desk          2000      nan                                              [blue]

And I want to change to result like this:

item_type  item_price  item_env                                        item_detail
sofa          2880     {natural wood, eco-friendly}                 {2pax,seglora}  
sofa          2400     {solid wood, recycle, reuse}                 {3pax,knisa}
chair         1200      nan                                               {Red}
desk          2000      nan                                              {blue}

I try to use replace and str.replace and some other method but it fail. So, I wanna ask how can do to get the expected result?

Answer 1

You need to escape the meta character of regular expressions, passing regex as True , whether you are using DataFrame.replace or Series.str.replace .

>>> df[['item_env', 'item_detail']] = (df[['item_env', 'item_detail']]
                                        .replace('\[', '{', regex=True)
                                        .replace('\]', '}', regex=True)
                                       )

OUTPUT:

  item_type  item_price                      item_env     item_detail
0      sofa        2880  {natural wood, eco-friendly}  {2pax,seglora}
1      sofa        2400  {solid wood, recycle, reuse}    {3pax,knisa}
2     chair        1200                           NaN           {Red}
3      desk        2000                           NaN          {blue}