Pandas——ix vs loc vs iloc区别

1. loc——通过行标签索引行数据

1.1 loc[1]表示索引的是第1行（index 是整数）

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = [0,1]

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.loc[1]

7. '''

8. a 4

9. b 5

10. c 6

11. '''

 

 

1.2 loc[‘d’]表示索引的是第’d’行（index 是字符）

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.loc['d']

7. '''

8. a 1

9. b 2

10. c 3

11. '''

 

 

1.3 如果想索引列数据，像这样做会报错

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.loc['a']

7. '''

8. KeyError: 'the label [a] is not in the [index]'

9. '''

 

 

1.4 loc可以获取多行数据

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.loc['d':]

7. '''

8. a b c

9. d 1 2 3

10. e 4 5 6

11. '''

 

 

1.5 loc扩展——索引某行某列

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.loc['d',['b','c']]

7. '''

8. b 2

9. c 3

10. '''

 

 

1,6 loc扩展——索引某列

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.loc[:,['c']]

7. '''

8. c

9. d 3

10. e 6

11. '''

 

 

当然获取某列数据最直接的方式是df.[列标签]，但是当列标签未知时可以通过这种方式获取列数据。

需要注意的是，dataframe的索引[1:3]是包含1,2,3的，与平时的不同。

2. iloc——通过行号获取行数据

2.1 想要获取哪一行就输入该行数字

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.loc[1]

7. '''

8. a 4

9. b 5

10. c 6

11. '''

 

 

2.2 通过行标签索引会报错

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.iloc['a']

7. '''

8. TypeError: cannot do label indexing on <class 'pandas.core.index.Index'> with these indexers [a] of <type 'str'>

9. '''

 

 

2.3 同样通过行号可以索引多行

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.iloc[0:]

7. '''

8. a b c

9. d 1 2 3

10. e 4 5 6

11. '''

 

 

2.4 iloc索引列数据

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.iloc[:,[1]]

7. '''

8. b

9. d 2

10. e 5

11. '''

 

 

3. ix——结合前两种的混合索引

3.1 通过行号索引

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.ix[1]

7. '''

8. a 4

9. b 5

10. c 6

11. '''

 

 

3.2 通过行标签索引

 

1. import pandas as pd

2. data = [[1,2,3],[4,5,6]]

3. index = ['d','e']

4. columns=['a','b','c']

5. df = pd.DataFrame(data=data, index=index, columns=columns)

6. print df.ix['e']

7. '''

8. a 4

9. b 5

10. c 6

11. '''