Question

我的数据框有2列Sector＆amp;部门名称，我想更新＆＃39;部门＆＃39; =＆＃39;部门名称＆＃39;其中＆＃39; Sector＆＃39;是空白

我有下面的脚本，但它的给出错误＆＃34; ValueError：无法从重复的轴重新索引＆＃34;

dataFinal.loc[dataFinal['Sector'] == '', 'Sector'] = \
           dataFinal.loc[dataFinal['Sector Name'] != '', 'Sector Name']

请帮忙。

Answer 1

我认为最好的是创建唯一索引，然后使用loc，mask或where来反转条件：

dataFinal = dataFinal.reset_index(drop=True)
#John Gals dol from comment 
dataFinal.loc[dataFinal['Sector'] == '', 'Sector'] = dataFinal['Sector Name']

或者：

m = dataFinal['Sector'] == ''
dataFinal['Sector'] = dataFinal['Sector'].mask(m, dataFinal['Sector Name'])

m = dataFinal['Sector'] != ''
dataFinal['Sector'] = dataFinal['Sector'].where(m, dataFinal['Sector Name'])

样品：

dataFinal = pd.DataFrame({'Sector':['a','ss',''],
                         'Sector Name':['r','t','y']}, index=[4,4,1])
print (dataFinal)
  Sector Sector Name
4      a           r
4     ss           t
1                  y

dataFinal = dataFinal.reset_index(drop=True)

m = dataFinal['Sector'] == ''
dataFinal['Sector'] = dataFinal['Sector'].mask(m, dataFinal['Sector Name'])
print (dataFinal)
  Sector Sector Name
0      a           r
1     ss           t
2      y           y

Answer 2

您可以使用np.where

dataFinal['Sector'] = \
    np.where(dataFinal['Sector'] == '', dataFinal['Sector Name'], dataFinal['Sector'])

感谢jezrael获取数据：

dataFinal

  Sector Sector Name
4      a           r
4     ss           t
1                  y

dataFinal['Sector'] = \
        np.where(dataFinal['Sector'] == '', dataFinal['Sector Name'], dataFinal['Sector'])
dataFinal
  Sector Sector Name
4      a           r
4     ss           t
1      y           y

Answer 3

您可以使用掩码查找数据框中Sector仅包含空格的所有行，然后使用此掩码应用相应的Sector Name：

mask = dataFinal['Sector'].str.isspace()
dataFinal.loc[mask, 'Sector'] = dataFinal.loc[mask, 'Sector Name']

如果行为空，则使用另一列中的值更新一列

3 个答案: