如何根据规则编辑python数据框的字符串类型列?

对于大学工作,我们在熊猫数据框中有一些世界位置的地理坐标:

      df = pd.DataFrame({'NAME': ['Paris', 'New York', 'Rio', 'Airport GRU', 'ORLY'],
               'GEO': ['POINT (2.31647 48.85)',
                       'POINT (-73.993457389558 40.731499671618)',
                       'POINT (-43.2 -22.9)',
                       'POINT (-46.47313507388693 -23.429382262746415)',
                       'GEOMETRYCOLLECTION EMPTY']})

      print(df)

        NAME                      GEO
       Paris                   POINT (2.31647 48.85)
       New York                POINT (-73.993457389558 40.731499671618)
        Rio                    POINT (-43.2 -22.9)
      Airport GRU              POINT (-46.47313507388693 -23.429382262746415)
        ORLY                   GEOMETRYCOLLECTION EMPTY

我想编辑“地理”专栏。起初,我想忽略“点”这个词,然后我想把它放在(纬度,经度)格式,因为这个列是点(经度-纬度)格式(它也是一个逗号)。

为了解决这个问题,我创建了两个单独的列来存储LAT和LONG(这部分正在工作):

       df2 = df.join(df['GEO'].str.extract(r'(?P<LONG>-?\d+\.\d+) (?P<LAT>-?\d+\.\d+)').astype(float))

      print(df2)

         NAME            GEO                                          LONG      LAT
        Paris       POINT (2.31647 48.85)                           2.316470   48.85000
        New York    POINT (-73.993457389558 40.731499671618)       -73.993457  40.731500
        Rio         POINT (-43.2 -22.9)                            -43.200000  -22.900000
        Airport GRU POINT (-46.47313507388693 -23.429382262746415) -46.473135  -23.429382
        ORLY        GEOMETRYCOLLECTION EMPTY                          NaN         NaN

但是,当我尝试创建一个新列来接收格式时:(LAT,LONG)。代码不起作用:

      df2['Result'] = "(" + df2['LAT'] + "," + df2['LONG'] + ")"

出现错误:“ufunctypererror:ufunc'add'不包含具有签名匹配类型的循环(dtype('<U1')、dtype('float64'))->无”

我希望输出为:

     NAME            GEO
   Paris            (48.85, 2.31647)
   New York         (40.731499671618, -73.993457389558)
   Rio              (-22.9, -43.2)
   Airport GRU      (-23.429382262746415, -46.47313507388693)
   ORLY             GEOMETRYCOLLECTION EMPTY]})