stack及unstack操作¶

In [ ]:

import pandas as pd
import numpy as np

import pandas as pd import numpy as np

In [ ]:

data = np.random.randint(0, 10, (2, 2))

data = np.random.randint(0, 10, (2, 2))

In [ ]:

data

data

Out[ ]:

array([[5, 2],
       [3, 5]])

In [ ]:

df1 = pd.DataFrame(data, index=['row1','row2'], columns=['col1','col2'])

df1 = pd.DataFrame(data, index=['row1','row2'], columns=['col1','col2'])

In [ ]:

df1

df1

Out[ ]:

	col1	col2
row1	5	2
row2	3	5

.stack()¶

In [ ]:

df1.stack()

df1.stack()

Out[ ]:

row1  col1    5
      col2    2
row2  col1    3
      col2    5
dtype: int64

In [ ]:

df1.stack().index

df1.stack().index

Out[ ]:

MultiIndex([('row1', 'col1'),
            ('row1', 'col2'),
            ('row2', 'col1'),
            ('row2', 'col2')],
           )

.unstack()¶

In [ ]:

df2 = df1.stack()

df2 = df1.stack()

In [ ]:

df2

df2

Out[ ]:

row1  col1    5
      col2    2
row2  col1    3
      col2    5
dtype: int64

In [ ]:

df2.unstack()

df2.unstack()

Out[ ]:

	col1	col2
row1	5	2
row2	3	5

In [ ]:

df2.unstack().index

df2.unstack().index

Out[ ]:

Index(['row1', 'row2'], dtype='object')