Question

当系统出现在“类型”列中时，我想删除该行中的所有值，但“名称”列中的值除外。当“类型”列中出现“硬件”时，我要删除该行中除“颜色”列中的值以外的所有值。之后，我想将“文本”列中所有不为空的单元格拆分为多行，并保留该列中为空的行。

这是我拥有的数据框：

df

  Type        Text                             Name        ID        Color
  System      aca\nmaca\nstream\nphase\n       Gary        123       Red
  System      aca\nmaca\nstream\nphase\n       Mary        3254      Yellow
  Hardware    a\nmaca\nstream\nphase\n         Jerry       158       White
  Software    ca\nmaca\nstream\nphase\n        Perry       56414     Green
  Software    aca\nmac\nstream\nphase\n        Jimmy       548       Blue
  System      aca\nmaca\nstream\nphase\n       Marc        5658      Black
  System      aca\nmaca\nstram\npha\n          John        867       Pink
  Hardware    aca\nma\nstream\nphase\n         Sam         665       Gray
  Hardware    aca\nmaca\nstream\nphase\n       Jury        5784      Azure
  System      aca\nmaca\nstream\nphase\n       Larry       5589      Fawn
  Software    aca\nmaca\nst\nphase\n           James       6568      Magenta
  System      aca\nmaca\nstream\nph\n          Kevin       568       Cyan

这是预期的结果：

  Type        Text                             Name        ID        Color
  System                                       Gary        
  System                                       Mary        
  Hardware                                                           White
  Software    ca                               Perry       56414     Green
  Software    maca                             Perry       56414     Green
  Software    stream                           Perry       56414     Green
  Software    phase                            Perry       56414     Green
  Software    aca                              Jimmy       548       Blue
  Software    mac                              Jimmy       548       Blue
  Software    stream                           Jimmy       548       Blue
  Software    phase                            Jimmy       548       Blue
  System                                       Marc         
  System                                       John        
  Hardware                                                           Gray
  Hardware                                                           Azure
  System                                       Larry       
  Software    aca                              James       6568      Magenta
  Software    maca                             James       6568      Magenta
  Software    st                               James       6568      Magenta
  Software    phase                            James       6568      Magenta
  System                                       Kevin

对于将单元格拆分为多行，我尝试了以下功能：

  def SepInRows(df, c):
      s = df[c].str.split('\n', expand=True).stack()
      i = s.index.get_level_values(0)
      df2 = df.loc[i].copy()
      df2[c] = s.values
      return df2

但是它在“文本”列中删除了带有空值的行，这不是我想要的。

该如何解决？

Answer 1

您可以在预处理中将mask与difference一起使用，然后this solution：

c1 = df.columns.difference(['Type','Name'])
c2 = df.columns.difference(['Type','Color'])

df[c1] = df[c1].mask(df['Type'] == 'System', np.nan)
df[c2] = df[c2].mask(df['Type'] == 'Hardware', np.nan)

cols = df.columns
df1 = (df.join(df.pop('Text').str.split('\n', expand=True)
              .stack()
              .reset_index(level=1, drop=True)
              .rename('Text'))
              ).reset_index(drop=True).reindex(columns=cols)

print (df1)
        Type    Text   Name       ID    Color
0     System     NaN   Gary      NaN      NaN
1     System     NaN   Mary      NaN      NaN
2   Hardware     NaN    NaN      NaN    White
3   Software      ca  Perry  56414.0    Green
4   Software    maca  Perry  56414.0    Green
5   Software  stream  Perry  56414.0    Green
6   Software   phase  Perry  56414.0    Green
7   Software          Perry  56414.0    Green
8   Software     aca  Jimmy    548.0     Blue
9   Software     mac  Jimmy    548.0     Blue
10  Software  stream  Jimmy    548.0     Blue
11  Software   phase  Jimmy    548.0     Blue
12  Software          Jimmy    548.0     Blue
13    System     NaN   Marc      NaN      NaN
14    System     NaN   John      NaN      NaN
15  Hardware     NaN    NaN      NaN     Gray
16  Hardware     NaN    NaN      NaN    Azure
17    System     NaN  Larry      NaN      NaN
18  Software     aca  James   6568.0  Magenta
19  Software    maca  James   6568.0  Magenta
20  Software      st  James   6568.0  Magenta
21  Software   phase  James   6568.0  Magenta
22  Software          James   6568.0  Magenta
23    System     NaN  Kevin      NaN      NaN

根据列中的值删除行中的值，然后在Pandas中将单元格拆分为多行

1 个答案: