正则表达式在熊猫中被分割

Question

你好，我有一个df，例如

def save_my_data(self, request):
  info['my_data']  = None
  serializer = serializers.MyModelSerializer(data=info)     
  # validate the data
  print( serializer.is_valid() ) #returns False
  print( serializer.errors ) #returns my_data: [This field can't be null]

我想删减最后一个COL1 NW_011625257.1_0 NW_011623521.1_1 NW_011623521.3_1 NW_011623521.4_1 NW_011623521.1 JZSA01007324.1_2 scaffold_1463_2 scaffold_1463 并得到

'_'

到目前为止，我已经尝试过：

COL1              COL2
NW_011625257.1    0
NW_011623521.1    1
NW_011623521.3    1
NW_011623521.4    1
NW_011623521.1    NaN 
JZSA01007324.1    2
scaffold_1463     2
scaffold_1463     NaN

相反，我得到了这样的输出：

df[['COL1','COL2']] = df.COL1.str.split(r'_(?!.*_)', expand=True)

这是我要选择的示例

Answer 1

您可以使用

df[['COL1','COL2']] = df.COL1.str.split(r"(?<=\d)_(?=\d+$)", expand=True)

请参见regex demo

模式详细信息：

• (?<=\d)-当前位置之前必须有一个数字
• _-下划线
• (?=\d+$)-当前位置的右边必须有1个以上的数字和字符串的结尾。

熊猫测试：

df = pd.DataFrame({'COL1': ['NW_011625257.1_0','NW_011623521.1_1','NW_011623521.3_1','NW_011623521.4_1','NW_011623521.1','JZSA01007324.1_2','scaffold_1463_2','scaffold_1463']})
>>> df[['COL2','COL3']] = df.COL1.str.split(r"(?<=\d)_(?=\d+$)", expand=True)
>>> df
               COL1            COL2  COL3
0  NW_011625257.1_0  NW_011625257.1     0
1  NW_011623521.1_1  NW_011623521.1     1
2  NW_011623521.3_1  NW_011623521.3     1
3  NW_011623521.4_1  NW_011623521.4     1
4    NW_011623521.1  NW_011623521.1  None
5  JZSA01007324.1_2  JZSA01007324.1     2
6   scaffold_1463_2   scaffold_1463     2
7     scaffold_1463   scaffold_1463  None